# Maestrix — Full Site Content for AI Agents > Generated: 2026-09-23T01:48:29.329Z > Site: https://maestrix.ai > Summary: https://maestrix.ai/llms.txt Maestrix is the AI marketing operating system. 100+ expert marketing agents & skills read from one persistent Marketing Brain to ship positioning, personas, content, ads, and GTM strategy in hours instead of weeks. --- # Page Directory ## Home URL: https://maestrix.ai/ Platform overview, audience-aware positioning, pricing snapshot. ## About URL: https://maestrix.ai/about Company story, mission, and the team behind Maestrix. ## Pricing URL: https://maestrix.ai/pricing Plans, credits, sprints, and top-up options. ## Platform URL: https://maestrix.ai/platform Full platform tour and capabilities. ## AI Marketing Agents URL: https://maestrix.ai/ai-marketing-agents The 100+ expert marketing agents & skills powering the platform. ## Marketing Sprints URL: https://maestrix.ai/ai-marketing-sprints 2 to 4 week done-for-you transformations. ## Services URL: https://maestrix.ai/services AI-Native agency services. ## Compare URL: https://maestrix.ai/compare Maestrix versus alternative tools and stacks. ## MCP URL: https://maestrix.ai/mcp Claude Cowork MCP integration. ## Free Marketing Skills URL: https://maestrix.ai/skills Library of 100+ free marketing skills you can run anywhere. ## FAQ URL: https://maestrix.ai/faq Common questions about plans, credits, and capabilities. ## Contact URL: https://maestrix.ai/contact Talk to a strategist or get support. ## Book a Demo URL: https://maestrix.ai/book-a-demo Schedule a personalized walkthrough. ## Marketing Strategy URL: https://maestrix.ai/solutions/marketing-strategy GTM, positioning, messaging. ## Competitive Research URL: https://maestrix.ai/solutions/competitive-research SWOT, battlecards, positioning maps. ## Search Ads URL: https://maestrix.ai/solutions/search-ads Google Ads campaign generation. ## Social Ads URL: https://maestrix.ai/solutions/social-ads Meta and LinkedIn campaign generation. ## Content Marketing URL: https://maestrix.ai/solutions/content-marketing Calendars, blogs, social, email. ## Customer Research URL: https://maestrix.ai/solutions/customer-research Personas, ICP sizing, journey maps. ## Marketing Brain URL: https://maestrix.ai/solutions/marketing-brain Persistent strategic memory layer. ## Agent Builder URL: https://maestrix.ai/solutions/agent-builder Build custom marketing agents. ## Sales Ops URL: https://maestrix.ai/solutions/sales-ops Sales enablement and outreach. ## What is a Buyer Persona URL: https://maestrix.ai/guides/what-is-a-buyer-persona ## What is Positioning and Messaging URL: https://maestrix.ai/guides/what-is-positioning-and-messaging ## What is Competitive Analysis URL: https://maestrix.ai/guides/what-is-competitive-analysis ## What is a Content Plan URL: https://maestrix.ai/guides/what-is-a-content-plan ## What is a Go-to-Market Strategy URL: https://maestrix.ai/guides/what-is-a-go-to-market-strategy ## What is Demand Generation URL: https://maestrix.ai/guides/what-is-demand-generation ## SaaS B2B Marketing Playbook URL: https://maestrix.ai/guides/saas-b2b-marketing-playbook ## Content and Influencer Marketing URL: https://maestrix.ai/guides/content-and-influencer-marketing ## Marketing Analytics and Automation URL: https://maestrix.ai/guides/marketing-analytics-and-automation ## GTM Stack Export Playbook URL: https://maestrix.ai/guides/gtm-stack-export-playbook ## 2026 Marketing AI Landscape URL: https://maestrix.ai/guides/2026-marketing-ai-landscape ## Skills Library Overview URL: https://maestrix.ai/guides/skills --- # Free Marketing Skills Each skill below is a complete, ready-to-run prompt. Source URL is provided so agents can cite the canonical location. --- ## Skill: Advertising Source: https://maestrix.ai/skills/advertising > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Advertising Professional-grade advertising tools covering paid search, social ads, and multi-platform creative. ## CONTEXT NEEDED - **Company/product**: What you sell and who buys it - **Campaign goal**: Awareness, lead gen, trial sign-ups, purchases - **Target audience**: Demographics, job titles, interests, or intent signals - **Platform**: Google, Meta, LinkedIn, TikTok, Reddit, or other - **Budget tier** *(optional)*: Starter / mid-market / enterprise - **Funnel stage** *(optional)*: TOFU / MOFU / BOFU --- ## Available Tools (3) ### 1. generate_google_ads Create complete Google Ads campaigns with RSAs, extensions, and keyword targeting. **Output includes:** - 15 headlines (≤30 chars each) - 4 descriptions (≤90 chars each) - Sitelink extensions - Callout extensions - Keyword recommendations - Negative keyword list ### 2. generate_meta_ads Create Facebook and Instagram ad campaigns with hooks and audience targeting. **Output includes:** - Primary text variations - Headlines and descriptions - Hook formulas for video - Audience targeting suggestions - Placement recommendations ### 3. create_ad_copy Generate versatile ad copy for any platform (LinkedIn, Twitter/X, TikTok, programmatic). --- ## Funnel Strategy **TOFU (Awareness):** - Broad interest targeting - Problem-aware messaging - Educational content **MOFU (Consideration):** - Retargeting segments - Solution-aware messaging - Social proof emphasis **BOFU (Conversion):** - High-intent keywords - Product-aware messaging - Strong CTAs and offers --- ## Output Rules - ✓ Character limits enforced per platform - ✓ Multiple copy variations provided - ✓ Funnel stage labeled on each asset - ✓ Platform specifications met - ✓ A/B testing recommendations included --- ## Example Prompts 1. **Google Ads**: "Create a Google Ads campaign targeting 'CRM software for small business'" 2. **Meta Ads**: "Generate Meta ads for our fitness app targeting busy professionals" 3. **Ad Copy**: "Write LinkedIn ad copy for our B2B recruiting platform" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this advertising plan to my {{companyName}} project" - **create_google_ads** (green, primary) — sendPrompt: "Create Google Ads RSA variants for {{companyName}} from this advertising plan" - **create_meta_ads** (neutral, secondary) — sendPrompt: "Create Meta ads for {{companyName}} that pair with the Google campaign" - **create_ad_copy** (neutral, secondary) — sendPrompt: "Generate ad copy variants for {{companyName}} to A/B test" - **create_landing_page** (neutral, secondary) — sendPrompt: "Build a landing page for {{companyName}} optimized for the advertising traffic" --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Analyze Competitor Source: https://maestrix.ai/skills/analyze-competitor > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create comprehensive competitor analysis with actionable strategic insights. ## CONTEXT NEEDED - Company name and description - Competitor name and URL - Your products/services - Your target audience - Specific focus areas (optional) ## COMPETITOR ANALYSIS FRAMEWORK ### 1. Company Overview **Basic Information** - Company name - Founded - Headquarters - Funding/Revenue (if available) - Employee count - Target market **Mission/Vision** What they say about themselves **Leadership** Key executives and their backgrounds ### 2. Product/Service Analysis **Product Portfolio** | Product | Description | Target Segment | Pricing | |---------|-------------|----------------|---------| **Feature Comparison** | Feature | Competitor | Your Product | Advantage | |---------|------------|--------------|-----------| **Unique Capabilities** What they can do that others can't **Gaps/Weaknesses** What's missing or underperforming ### 3. Positioning & Messaging **Positioning Statement** (Inferred) For [target], [Company] is the [category] that [key benefit] because [reason]. **Value Propositions** | Claim | Proof Point | Strength | |-------|-------------|----------| **Messaging Themes** - Primary message - Supporting messages - Emotional appeals - Rational appeals **Tagline/Slogan** Analysis of effectiveness ### 4. Pricing Strategy **Pricing Model** - Structure (subscription, one-time, usage) - Tiers and features - Comparison to market **Promotional Tactics** - Discounts - Free trials - Freemium **Price Positioning** Premium / Mid-market / Budget ### 5. Go-to-Market Analysis **Sales Model** - Self-serve vs. sales-led - Sales team structure - Channel partners **Marketing Channels** | Channel | Investment Level | Effectiveness | |---------|------------------|---------------| **Content Strategy** - Blog topics and frequency - Resource types - SEO focus areas **Social Media Presence** | Platform | Followers | Engagement | Content Type | |----------|-----------|------------|--------------| ### 6. Customer Analysis **Target Segments** Who they're selling to **Customer Proof** - Logos - Case studies - Testimonials **Customer Sentiment** - Review analysis (G2, Capterra, etc.) - Common praise - Common complaints ### 7. SWOT Analysis **Strengths** - [List with evidence] **Weaknesses** - [List with evidence] **Opportunities** - [Market/strategic opportunities] **Threats** - [Competitive/market threats] ### 8. Competitive Battlecard **Quick Facts** - Founded: - Funding: - Customers: - Key products: **When We Win** Scenarios where we outperform **When We Lose** Scenarios where they outperform **Objection Responses** | Their Claim | Our Response | |-------------|--------------| **Landmines** Questions to ask that highlight their weaknesses ### 9. Strategic Recommendations **How to Differentiate** - Message positioning - Feature emphasis - Segment focus **How to Compete** - Direct attacks - Flanking strategies - Market education **How to Monitor** - Key signals to track - Review frequency ### 10. Competitive Intelligence Sources **Public Sources** - Website changes (archive.org) - Job postings (growth areas) - Press releases - Patent filings - SEC filings (if public) **Community Sources** - Review sites - Social mentions - Community forums - Industry events ## QUALITY RULES 1. NO opening comments 2. Evidence-based analysis 3. Actionable recommendations 4. Regular update cadence ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C **Layout spec:** Header row (one card per competitor with score badge) → feature matrix (color-coded: --success win, --destructive lose, --muted neutral) → battle card block (positioning, weakness, counter) → bottom highlights (--warning gap, --success opportunity). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Battle card ↗' - sendPrompt: 'Create a competitive battle card for {{companyName}} — {{description}}, based on this competitor analysis' Secondary actions (neutral styling): - Label: 'Counter-messaging ↗' sendPrompt: 'Create counter-messaging for {{companyName}} — {{description}}, addressing the competitor weaknesses identified' - Label: 'Positioning update ↗' sendPrompt: 'Update the positioning for {{companyName}} — {{description}}, incorporating the competitive gaps we found' - Label: 'Cold email using gaps ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, leveraging the competitive advantages from this analysis' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Analyze Content Gaps Source: https://maestrix.ai/skills/analyze-content-gaps > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Content Gap Analysis Reverse-engineer a competitor's content strategy, identify topics they're not covering (or covering poorly), map keyword clusters worth owning, and produce a prioritized 12-week content calendar. ## CONTEXT NEEDED - **Competitor URL**: The website to analyze (required) - **Your company/product**: What you sell and who buys it - **Your existing content** *(optional)*: Topics you already cover - **Target audience** *(optional)*: Roles, industries, or buyer personas - **Priority themes** *(optional)*: Topics or keywords you care about most --- ## ANALYSIS FRAMEWORK ### 1. Competitor Content Audit Inventory what the competitor publishes, how deep they go, and where their content clusters. | Topic Cluster | # Pages | Content Types | Estimated Depth | Recency | |---------------|---------|---------------|-----------------|---------| | (e.g., "Onboarding best practices") | 12 | Blog, Guide, Video | Deep — pillar + 8 supporting | Last 30 days | **Content types observed**: Blog posts, guides, ebooks, webinars, case studies, videos, podcasts, templates, tools **Publishing cadence**: Estimated frequency and consistency **Top-performing signals**: Pages that appear well-optimized (H1 structure, internal linking, schema) ### 2. Gap Analysis Matrix Topics and keyword clusters the competitor is NOT covering — or covering weakly — that represent opportunity for you. | Gap Topic | Why It's a Gap | Search Intent | Difficulty | Opportunity Score | Your Advantage | |-----------|---------------|---------------|------------|-------------------|----------------| | (topic) | Not covered / thin / outdated | Informational / Commercial / Transactional | Low-Med-High | 1-10 | Why you can own this | **Scoring methodology**: - **Opportunity Score** = (Search demand × Relevance to your ICP) ÷ Competitive difficulty - Prioritize gaps where you have subject-matter authority or product differentiation ### 3. Keyword Cluster Map Group identified opportunities by search intent. **Informational** (top-of-funnel awareness) | Cluster | Primary Keyword | Supporting Keywords | Monthly Volume Est. | Current Gap | |---------|----------------|--------------------|--------------------|-------------| **Commercial** (mid-funnel comparison/evaluation) | Cluster | Primary Keyword | Supporting Keywords | Monthly Volume Est. | Current Gap | |---------|----------------|--------------------|--------------------|-------------| **Transactional** (bottom-funnel purchase intent) | Cluster | Primary Keyword | Supporting Keywords | Monthly Volume Est. | Current Gap | |---------|----------------|--------------------|--------------------|-------------| ### 4. Content Calendar Scaffold (12 Weeks) Prioritized by gap opportunity score. Each entry includes format recommendation. | Week | Topic | Format | Target Keyword Cluster | Intent Stage | Effort | Expected Impact | |------|-------|--------|----------------------|--------------|--------|-----------------| | 1 | | Blog post / Pillar page / Guide / Video / Tool | | TOFU/MOFU/BOFU | S/M/L | | | 2 | | | | | | | | ... | | | | | | | | 12 | | | | | | | **Format recommendations**: - **Quick wins** (weeks 1-4): Blog posts targeting low-difficulty gaps - **Authority builders** (weeks 5-8): Pillar pages, comprehensive guides - **Conversion drivers** (weeks 9-12): Comparison pages, case studies, tools ### 5. Quick Wins Low-effort, high-impact content pieces to publish first (before the 12-week calendar). | Quick Win | Why Now | Format | Time to Publish | Expected Result | |-----------|---------|--------|-----------------|-----------------| | | Competitor has nothing here + high search demand | | 1-3 days | | ### 6. Internal Linking Strategy How to connect new gap content to your existing pages for maximum SEO lift. | New Content | Links To (Existing) | Links From (Existing) | Anchor Text Suggestion | |-------------|--------------------|-----------------------|----------------------| --- ## Output Rules - ✓ No opening comments or fluff - ✓ Direct substantive analysis based on actual scraped content - ✓ Opportunity scores must be justified, not arbitrary - ✓ Calendar entries must be specific topics, not generic categories - ✓ Include "why you can win" rationale for each gap - ✓ All keyword suggestions must be realistic and intent-matched --- ## Example Prompts 1. "Analyze content gaps vs hubspot.com for our CRM startup" 2. "What topics is intercom.com missing that we should own?" 3. "Find content gaps between us and zendesk.com in the support space" 4. "Reverse-engineer lemlist.com's content strategy and find opportunities" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full analysis output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (competitor URL + opportunity count) → tab bar (5 tabs: Audit, Gaps, Keywords, Calendar, Quick Wins) → per-tab sub-layouts: audit (card list with depth indicators), gaps (scored card list with opportunity bars), keywords (grouped by intent with volume chips), calendar (week-by-week timeline cards), quick wins (priority card list) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ### What's next actions | Position | Label | Prompt | |----------|-------|--------| | Primary | 💾 Save to project ↗ | Save this content gap analysis to my {{companyName}} project | | Secondary 1 | Content plan from gaps ↗ | Create a content plan for {{companyName}} based on the content gaps we just identified | | Secondary 2 | SEO strategy ↗ | Create an SEO strategy for {{companyName}} targeting the keyword clusters from this gap analysis | | Secondary 3 | Write first article ↗ | Write a blog article for {{companyName}} on the #1 quick win topic from our gap analysis | --- ## Skill: Analyze Homepage Source: https://maestrix.ai/skills/analyze-homepage > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Analyze homepage for conversion optimization with actionable improvement recommendations. ## CONTEXT NEEDED - Company name and description - Homepage URL - Target audience - Primary conversion goal - Current conversion rate (optional) ## HOMEPAGE ANALYSIS FRAMEWORK ### 1. LIFT Model Scoring Score each factor on a 1-10 scale: | LIFT Factor | Score | Assessment | Priority Fixes | |-------------|-------|------------|----------------| | Value Proposition | /10 | Is unique value immediately clear? | | | Relevance | /10 | Does it match visitor expectations? | | | Clarity | /10 | Is information easy to understand? | | | Anxiety | /10 | Are trust concerns addressed? | | | Distraction | /10 | Is focus on primary CTA? | | | Urgency | /10 | Is there motivation to act now? | | | **Total LIFT Score** | /60 | | | ### 2. First Impression Audit (5-Second Test) **Above-the-Fold Assessment** - What's immediately clear? - What's confusing? - Is the value proposition visible? - Is the CTA prominent? - Does it match user expectations? **Score**: [1-10] **Priority Issues**: [List] ### 2. Value Proposition Analysis **Current Value Prop** - Where it appears - Clarity score (1-10) - Differentiation score (1-10) - Believability score (1-10) **Recommended Improvements** - Headline rewrite - Subhead rewrite - Supporting proof point ### 3. Navigation & Information Architecture **Navigation Assessment** - Menu clarity - Link labels - Path to key pages - Mobile navigation **Issues Found** | Issue | Location | Impact | Recommendation | |-------|----------|--------|----------------| ### 4. CTA Analysis **Primary CTA** - Location - Visibility - Copy - Design contrast - Action clarity **Secondary CTAs** - Hierarchy - Competing actions - User journey mapping **Recommendations** - CTA copy alternatives - Placement improvements - Design suggestions ### 5. Trust Signals Audit | Trust Element | Present? | Location | Effectiveness | Improvement | |---------------|----------|----------|---------------|-------------| | Customer logos | | | | | | Testimonials | | | | | | Case studies | | | | | | Security badges | | | | | | Ratings/reviews | | | | | | Certifications | | | | | | Social proof | | | | | | Guarantees | | | | | ### 6. Content Hierarchy Analysis **Section-by-Section Review** | Section | Purpose | Effectiveness | Issues | Recommendations | |---------|---------|---------------|--------|-----------------| ### 7. Messaging Clarity **Headline Analysis** - Current headline - Clarity score - Emotional impact - Improvement options **Body Copy Analysis** - Readability score - Benefit clarity - Jargon issues - Scanability ### 8. Visual Design Assessment **Design Elements** - Visual hierarchy - White space usage - Color effectiveness - Image relevance - Mobile responsiveness **Issues & Recommendations** | Element | Issue | Recommendation | |---------|-------|----------------| ### 9. Technical Performance **Speed Assessment** - Load time issues - Image optimization - Script concerns **SEO Elements** - Title tag - Meta description - H1 structure - Schema markup ### 10. Competitive Comparison | Element | Your Site | Competitor A | Competitor B | Gap | |---------|-----------|--------------|--------------|-----| ### 11. Prioritized Recommendations **Quick Wins** (Implement immediately) 1. [Change] - [Expected impact] 2. ... **Medium Effort** (1-2 weeks) 1. [Change] - [Expected impact] 2. ... **Strategic Changes** (Redesign elements) 1. [Change] - [Expected impact] 2. ... ### 12. Form Friction Analysis (if applicable) **Form Assessment** | Field | Necessary? | Friction Level | Optimization | |-------|------------|----------------|--------------| **Form Metrics** - Total fields: [N] - Required fields: [N] - Multi-step potential: [Yes/No] - Inline validation: [Yes/No] - Error handling: [Good/Poor] **Recommendations** - [ ] Field reduction opportunities - [ ] Multi-step conversion - [ ] Inline validation improvements ### 13. Micro-Conversion Analysis **Secondary Conversion Paths** | Micro-Conversion | Current State | Importance | Optimization | |------------------|---------------|------------|--------------| | Email capture | | | | | Content download | | | | | Video engagement | | | | | Chat initiation | | | | | Feature exploration | | | | ### 14. A/B Test Roadmap | Test | Hypothesis | Priority | Expected Lift | |------|------------|----------|---------------| ## QUALITY RULES 1. NO opening comments 2. Include LIFT model scoring with total 3. Specific, actionable recommendations 4. Prioritized by impact 5. Include before/after examples 6. Assess form friction if forms present ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** H **Layout spec:** Score header (large number, color-coded: >80 --success, 60-80 --warning, <60 --destructive, sub-score chips) → ranked recommendation list (severity badge, title, effort/impact chips) → quick wins (--success tint) → implementation timeline (Now/30d/90d). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Page CRO ↗' - sendPrompt: 'Run a CRO audit for {{companyName}} — {{description}}, diving deeper into the conversion issues found in this homepage analysis' Secondary actions (neutral styling): - Label: 'Signup flow CRO ↗' sendPrompt: 'Analyze the signup flow for {{companyName}} — {{description}}, building on the homepage audit findings' - Label: 'Landing page rewrite ↗' sendPrompt: 'Rewrite the landing page for {{companyName}} — {{description}}, fixing the issues identified in this analysis' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, based on the optimization recommendations from this audit' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Analyze Page Cro Source: https://maestrix.ai/skills/analyze-page-cro > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Analyze any marketing page for conversion rate optimization using the LIFT Model and behavioral psychology frameworks. ## CONTEXT NEEDED - Page URL to analyze - Page type (landing, pricing, feature, product, signup) - Primary conversion goal - Target audience - Current conversion rate (optional) ## LIFT MODEL ANALYSIS FRAMEWORK ### 1. Value Proposition Assessment **Clarity Score**: [1-10] | Element | Current State | Issues | Recommendations | |---------|--------------|--------|-----------------| | Headline | | | | | Subheadline | | | | | Hero section | | | | | Primary benefit | | | | **Value Communication Analysis** - Is the unique value immediately clear? - Does it answer "Why should I choose you?" - Is the benefit quantified where possible? - Does it match visitor intent? **Recommendations** - [ ] Headline rewrite with benefit focus - [ ] Subhead supporting the primary claim - [ ] Proof point to back the promise ### 2. Relevance Analysis **Message Match Score**: [1-10] | Traffic Source | Expected Message | Page Message | Match? | |----------------|------------------|--------------|--------| | Paid Search | | | | | Social Ads | | | | | Email | | | | | Organic | | | | **Audience Alignment** - Persona match: [1-10] - Language match: [1-10] - Visual match: [1-10] - Offer match: [1-10] **Recommendations** - [ ] Message match improvements - [ ] Audience-specific content blocks - [ ] Dynamic personalization opportunities ### 3. Clarity Optimization **Information Architecture** | Section | Purpose Clear? | Action Clear? | Improvements | |---------|---------------|---------------|--------------| | Above fold | | | | | Features | | | | | Benefits | | | | | Social proof | | | | | CTA areas | | | | **Cognitive Load Analysis** - Form field reduction opportunities - Copy simplification needs - Visual hierarchy improvements - Navigation streamlining **Recommendations** - [ ] Simplify primary message - [ ] Reduce decision complexity - [ ] Improve visual hierarchy ### 4. Anxiety Reduction **Trust Signals Audit** | Trust Element | Present | Location | Effectiveness | Priority Fix | |---------------|---------|----------|---------------|--------------| | Security badges | | | | | | Customer logos | | | | | | Testimonials | | | | | | Guarantees | | | | | | Privacy policy | | | | | | Contact info | | | | | | Third-party reviews | | | | | **Risk Perception Analysis** - What concerns might visitors have? - Where do they need reassurance? - What objections aren't addressed? **Recommendations** - [ ] Add missing trust elements - [ ] Reposition existing proof - [ ] Address top objections ### 5. Distraction Elimination **Attention Audit** | Distraction | Type | Location | Impact | Action | |-------------|------|----------|--------|--------| | | Navigation | | | | | | Competing CTAs | | | | | | Unrelated content | | | | | | Exit links | | | | **Focus Analysis** - Attention ratio (links:CTAs): [X:1] - Competing actions count: [N] - Visual distraction score: [1-10] **Recommendations** - [ ] Remove/minimize distractions - [ ] Single-focus CTA hierarchy - [ ] Simplify navigation ### 6. Urgency & Motivation **Motivation Elements** | Element | Current | Effectiveness | Improvement | |---------|---------|---------------|-------------| | Scarcity signals | | | | | Time limits | | | | | Social proof velocity | | | | | Loss aversion triggers | | | | | Benefit reminders | | | | **Psychological Triggers Assessment** - FOMO elements: [Present/Missing] - Social proof freshness: [Current/Stale] - Benefit reinforcement: [Strong/Weak] **Recommendations** - [ ] Add authentic urgency - [ ] Implement social proof velocity - [ ] Reinforce benefits near CTA ## FORM FRICTION ANALYSIS ### Form Assessment (if applicable) | Field | Necessary? | Friction Level | Optimization | |-------|------------|----------------|--------------| | | | | | **Form Metrics** - Total fields: [N] - Required fields: [N] - Multi-step potential: [Yes/No] - Inline validation: [Yes/No] - Error handling: [Good/Poor] **Recommendations** - [ ] Field reduction opportunities - [ ] Multi-step conversion - [ ] Inline validation additions - [ ] Error message improvements ## MICRO-CONVERSION ANALYSIS ### Secondary Actions Audit | Micro-Conversion | Current | Importance | Optimization | |------------------|---------|------------|--------------| | Email capture | | | | | Content download | | | | | Video engagement | | | | | Feature exploration | | | | | Chat initiation | | | | **Funnel Pathway Analysis** - Primary path to conversion - Alternative paths available - Recovery mechanisms for abandonment ## CRO SCORING SUMMARY | LIFT Factor | Score (1-10) | Priority | Quick Win Available | |-------------|--------------|----------|---------------------| | Value Proposition | | | | | Relevance | | | | | Clarity | | | | | Anxiety | | | | | Distraction | | | | | Urgency | | | | | **Overall CRO Score** | **/60** | | | ## PRIORITIZED RECOMMENDATIONS ### Quick Wins (Immediate) | Change | Expected Impact | Effort | Priority | |--------|-----------------|--------|----------| | | | | | ### Medium Effort (1-2 weeks) | Change | Expected Impact | Effort | Priority | |--------|-----------------|--------|----------| | | | | | ### Strategic Changes (Redesign) | Change | Expected Impact | Effort | Priority | |--------|-----------------|--------|----------| | | | | | ## A/B TEST ROADMAP | Test | Hypothesis | LIFT Factor | Priority | Expected Lift | |------|------------|-------------|----------|---------------| | | | | | | ## QUALITY RULES 1. NO opening comments - start directly with analysis 2. Score every LIFT factor 1-10 3. Provide specific, actionable recommendations 4. Include before/after examples where possible 5. Prioritize by impact and effort ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** H **Layout spec:** Score header (large number, color-coded: >80 --success, 60-80 --warning, <60 --destructive, sub-score chips) → ranked recommendation list (severity badge, title, effort/impact chips) → quick wins (--success tint) → implementation timeline (Now/30d/90d). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'A/B testing plan ↗' - sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, based on the CRO issues identified' Secondary actions (neutral styling): - Label: 'Landing page rewrite ↗' sendPrompt: 'Rewrite the landing page for {{companyName}} — {{description}}, fixing the conversion issues found' - Label: 'Signup flow CRO ↗' sendPrompt: 'Analyze the signup flow for {{companyName}} — {{description}}, continuing the CRO audit' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, with improved landing page messaging' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Analyze Reddit Opportunity Source: https://maestrix.ai/skills/analyze-reddit-opportunity > **EXECUTION RULE**: Generate the complete subreddit analysis yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Given a product URL or description, identify the best subreddits for organic marketing, score audience fit and self-promotion friction, and lay out a phased entry strategy that won't get the user banned. ## CONTEXT NEEDED Before generating, the user should provide: - **Product / brand name**: e.g. "Maestrix" - **One-liner**: what it does in one sentence - **Target audience**: who you're trying to reach (titles, behaviors, interests) - **URL** *(optional)*: product website for deeper audience inference - **Goals**: brand awareness / traffic / leads / community building - **Existing Reddit account age + karma** *(optional)*: gates which subreddits are viable ## THE 5-SECTION FRAMEWORK ### 1. Top Subreddits — Ranked Generate a table with 10-12 subreddit recommendations. Columns: | Rank | Subreddit | Subscribers | Activity Level | Audience Fit /10 | Self-Promo Rules | ROI Score (★★★★★) | For EACH subreddit row, also include below the table a deep card with: - **Why it fits**: 2-sentence explanation of audience overlap - **Post types that work**: story / question / data drop / AMA / discussion + which performs best - **Self-promotion rules**: what's allowed, what gets you banned (cite the rule) - **Best posting times**: day + time window (specify timezone, default EST) - **Flair requirements**: required flairs for posts to surface - **Account requirements**: minimum age + karma typical mods enforce - **Top mod sensitivities**: 1-2 cultural landmines ### 2. Audience Overlap Analysis Map the product's target personas to subreddit demographics. - **Persona profile per top 3 subreddits**: age range, professional level, pain points discussed in last 30 days - **Common questions asked**: 5 recurring questions you can answer - **Competitor mentions**: which competitors get named + how often + sentiment - **Vocabulary the community uses**: 10 in-group terms to mirror (or avoid) ### 3. Competitor Presence For each of the top 3 competitors: | Competitor | Active In | Post Frequency | Engagement Level | Strategy Type | Lessons | Followed by: 3 specific tactics a competitor uses that work + 2 that visibly fail. ### 4. Phased Entry Strategy Sequential plan, no shortcuts. - **Week 1-2 — Comment-only phase**: target 30 substantive comments across top 5 subreddits, no links - **Week 3-4 — First value posts**: 2 data drops or insight posts, still no product mention - **Month 2 — Story posts with subtle context**: 4 posts, brand mention only when context demands - **Month 3+ — Established presence**: AMAs, community engagement, structured value posts - **Per phase**: success metrics (karma gain, comment ratio, mod warnings), kill criteria, escalation rules ### 5. Risk Assessment - **Subreddits to AVOID** (3 minimum): name + why (mod history, anti-vendor culture, brigading risk) - **Common mistakes that get you banned**: 6 specific behaviors - **Shadowban indicators**: how to detect + how to recover - **Account age / karma requirements**: minimums by tier of subreddit - **Spam filter triggers**: link domains, posting cadence, account age signals ## QUALITY RULES 1. NO opening commentary — start with the subreddit table 2. Every subreddit recommendation includes its actual self-promotion policy, not a generalization 3. ROI scores justified with specific reasoning (subscribers × activity × fit × promo friction) 4. Minimum 10 subreddit recommendations 5. Phased entry strategy never suggests launching with promotional posts 6. All posting times specify timezone 7. Audience vocabulary list uses real community terminology, not corporate paraphrase --- ## UI Rendering Contract **Pattern:** B (Tabbed Dashboard) **Layout spec:** Persistent header (brand name + "Reddit Opportunity Map" badge + total subreddit count + audience reach estimate as stat chips). Tab bar with 5 tabs: **Top Subreddits** / **Audience Overlap** / **Competitor Presence** / **Entry Strategy** / **Risks**. Top Subreddits tab default — renders the ranked table with expandable rows that reveal the per-subreddit deep card. Entry Strategy tab renders as a 4-phase timeline with connector lines. Risk tab uses destructive/0.1 backgrounds for the AVOID list. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this Reddit opportunity map to my {{companyName}} project" - **create_reddit_posts** (green, primary) — sendPrompt: "Create Reddit posts for {{companyName}} targeting the top-ranked subreddits from this analysis" - **find_reddit_threads** (neutral, secondary) — sendPrompt: "Find active Reddit threads for {{companyName}} to engage in this week" - **create_content_plan** (neutral, secondary) — sendPrompt: "Create a content plan for {{companyName}} with Reddit distribution built into the calendar" - **analyze_competitor** (neutral, secondary) — sendPrompt: "Deeper competitor analysis for {{companyName}} on Reddit — what's working, what's not" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Analyze Youtube Trends Source: https://maestrix.ai/skills/analyze-youtube-trends > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Analyze YouTube trends and generate actionable video content opportunities. ## CONTEXT NEEDED - Company name and description - Industry/niche focus - Target audience on YouTube - Current YouTube presence (optional) - Competitor channels (optional) ## YOUTUBE TRENDS FRAMEWORK ### 1. Niche Trend Overview **Trending Topic Categories** | Topic | Trend Direction | Search Volume | Competition | Opportunity Score | |-------|-----------------|---------------|-------------|-------------------| Identify 10-15 relevant trending topics. **Seasonal Patterns** - Q1 trends - Q2 trends - Q3 trends - Q4 trends ### 2. Content Format Analysis **Performing Formats in Niche** | Format | Avg Views | Avg Engagement | Best For | Difficulty | |--------|-----------|----------------|----------|------------| | Long-form tutorial | | | Deep learning | | | Short explainer | | | Quick tips | | | List/compilation | | | Discovery | | | Interview/podcast | | | Authority | | | Behind-the-scenes | | | Connection | | | Reaction/commentary | | | Engagement | | | Comparison/versus | | | Decisions | | | Challenge/trend | | | Virality | | ### 3. Competitor Channel Analysis For each competitor (3-5): - Channel name - Subscriber count - Upload frequency - Top-performing videos (5) - Content themes - Thumbnail style - Title patterns - Gaps/opportunities ### 4. Video Ideas (20 Concepts) For each idea: **Title**: [Click-worthy title under 60 chars] **Hook**: [First 5 seconds script] **Format**: [Video type] **Length**: [Target duration] **Thumbnail Concept**: [Visual description] **Target Keyword**: [Search term] **Estimated Potential**: [Views projection] **Difficulty**: [Production complexity] ### 5. Shorts Strategy **Trending Shorts Formats** - Format types working now - Audio/music trends - Hook patterns - CTA approaches **Shorts Ideas** (10) - Concept, hook, CTA ### 6. Title & Thumbnail Patterns **High-Performing Title Formulas** - Formula 1: [Template + example] - Formula 2: [Template + example] - Formula 3: [Template + example] - Formula 4: [Template + example] - Formula 5: [Template + example] **Thumbnail Best Practices** - Text usage - Face/emotion - Colors - Contrast - Before/after layout ### 7. Keyword Opportunities | Keyword | Monthly Searches | Competition | Content Angle | |---------|------------------|-------------|---------------| List 20 keyword opportunities. ### 8. Content Calendar Template | Week | Main Video | Topic | Shorts (3) | Community Post | |------|------------|-------|------------|----------------| | 1 | | | | | | 2 | | | | | | 3 | | | | | | 4 | | | | | ### 9. Engagement Tactics **Community Building** - Comment strategies - Polls and community posts - Live stream opportunities - Collaboration ideas **Algorithm Signals** - Watch time optimization - CTR improvement tactics - Session time strategies ### 10. Measurement Framework | Metric | Target | Tracking Method | |--------|--------|-----------------| | Subscriber growth | | | | Views per video | | | | Watch time (hours) | | | | CTR | | | | Engagement rate | | | ## QUALITY RULES 1. NO opening comments 2. Data-informed recommendations 3. Specific, actionable video concepts 4. Platform-native strategies ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** G **Layout spec:** Filter bar (category chips, JS toggle) → group cards (one per category) → items as chips (--muted bg, hover --accent) → prominence tier (top items --primary/0.12 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Video script ↗' - sendPrompt: 'Create a video script for {{companyName}} — {{description}}, based on the trending topics discovered' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, incorporating video trends' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, aligned with these YouTube trends' - Label: 'Campaign ideas ↗' sendPrompt: 'Generate campaign ideas for {{companyName}} — {{description}}, leveraging these trend insights' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Business Foundations Source: https://maestrix.ai/skills/business-foundations > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Business Foundations Strategic marketing foundation tools for building personas, positioning, messaging, value propositions, and brand voice. ## CONTEXT NEEDED - **Company/product**: What you sell, who buys it, and why they buy it - **Market context**: B2B / B2C / D2C, industry, company stage - **Target audience**: Roles, segments, or demographics to focus on - **Competitive landscape** *(optional)*: Key competitors and your differentiation - **Existing brand voice samples** *(optional)*: For brand voice tool only --- ## Available Tools (5) ### 1. generate_persona Create comprehensive buyer personas with 10 deep sections including professional profile, motivations, pain points, and decision journey. ### 2. create_positioning Generate strategic positioning canvas with market category, competitive alternatives, and SWOT analysis. ### 3. create_messaging Build 13-section messaging framework with core value proposition and objection handling. ### 4. generate_value_proposition Create value proposition ladders with feature-benefit-outcome chains. ### 5. create_brand_voice Define brand voice guidelines with tone, vocabulary, and channel modulation rules. --- ## Output Rules - ✓ No opening comments or fluff - ✓ Direct substantive content - ✓ 400+ lines of structured output - ✓ Auto-inference of B2B/B2C/D2C context - ✓ Credible market reasoning, no filler --- ## Example Prompts 1. **Persona**: "Create a buyer persona for enterprise CFOs evaluating finance software" 2. **Positioning**: "Generate positioning for our CRM competing with Salesforce and HubSpot" 3. **Messaging**: "Build a messaging framework for our HR tech platform" 4. **Value Prop**: "Create value proposition ladders for our AI writing assistant" 5. **Brand Voice**: "Define brand voice guidelines for our fintech startup" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save these business foundations to my {{companyName}} project" - **run_marketing_foundation** (green, primary) — sendPrompt: "Run the full marketing foundation for {{companyName}} — persona, positioning, messaging, value prop, brand voice" - **create_positioning** (neutral, secondary) — sendPrompt: "Deepen the positioning canvas for {{companyName}}" - **generate_persona** (neutral, secondary) — sendPrompt: "Generate a complete buyer persona for {{companyName}}" - **create_messaging** (neutral, secondary) — sendPrompt: "Build a messaging framework for {{companyName}}" --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Choose Positioning Strategy Source: https://maestrix.ai/skills/choose-positioning-strategy > **EXECUTION RULE**: Generate the complete strategy comparison yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Pick the right Fletch positioning strategy for the company, then route to the matching downstream skill. ## CONTEXT NEEDED - Company name and description - Target audience - Key competitors (optional) - Current category, if any (optional) ## THE 3 STRATEGIES 1. **Mature Category** — Established category, known leaders. Compete on differentiation. 2. **Immature Category** — Category exists but fragmented. Compete by clarifying it. 3. **New Category** — No category. Compete by naming the problem and owning the frame. ## STRATEGY COMPARISON MATRIX 8 rows × 3 columns (Mature | Immature | New): 1. Market Awareness 2. Category Maturity 3. Competition Level 4. Buyer Behavior 5. Risk Profile 6. Time to Traction 7. Messaging Approach 8. Recommended Next Skill Each cell is one sentence specific to the company. No boilerplate. ## STRATEGY SCORING For each of the 3 strategies: - Fit Score (0–10) - Why (2 sentences) - Risk (1 sentence) ## RECOMMENDED STRATEGY - Chosen Strategy - Rationale (3 sentences) - What this unlocks (2 bullets) - What to avoid (1 bullet) ## QUALITY RULES 1. NO opening comments 2. Be specific to the company, not theoretical 3. The matrix IS the deliverable 4. Pick one winner — no fence-sitting ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full output above, render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C (Comparison Grid) **Layout spec:** Header card with company name + chosen strategy badge (--primary background) → 3-column comparison grid (Mature | Immature | New) with 8 labeled rows, sticky row labels on the left, recommended column highlighted with `rgba(31,157,85,0.08)` background and 2px solid `#1F9D55` left border → 3 score cards below the matrix (one per strategy, with 0–10 score chip, why, risk) → Recommendation block (primary-tinted card with chosen strategy, rationale, unlocks, avoid). **Design system — hardcoded values:** | Token | Value | |---|---| | background | #ffffff | | foreground | #222832 | | card | #fafafa | | primary | #1F9D55 | | primary-foreground | #ffffff | | muted | #f2f3f5 | | muted-foreground | #717d8a | | border | #e2e4e8 | | warning | #e8a307 | | destructive | #ef4444 | | font | Inter, system-ui, -apple-system, sans-serif | Border radius: 8px chips, 12px cards. Weights 400 and 500 only. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button. Pre-fill with company name, description, and the chosen strategy from this output so the next skill is informed by the verdict. Primary action (green styling) — route based on the chosen strategy: - If chosen strategy is **Mature Category** or **Immature Category**: - Label: 'Build positioning ↗' - sendPrompt: 'Create positioning for {{companyName}} — {{description}}. Use the {{chosenStrategy}} strategy I just selected. Frame of reference, key differentiator, and proof points should reflect that choice.' - If chosen strategy is **New Category**: - Label: 'Design new category ↗' - sendPrompt: 'Create category positioning for {{companyName}} — {{description}}. I confirmed the New Category strategy. Define the category name, manifesto, and lightning strike plan.' Secondary actions (neutral styling): - Label: 'Analyze competitors ↗' sendPrompt: 'Analyze the competitive landscape for {{companyName}} — {{description}} before I finalize positioning' - Label: 'Define ICP ↗' sendPrompt: 'Define the ICP and sizing for {{companyName}} — {{description}} so positioning targets the right buyer' **Rules:** - Generate complete content first, then render - Never make a secondary API call - Zero CSS variables — hardcoded values only - Attribution footer on every widget --- ## Skill: Complete Bundle Source: https://maestrix.ai/skills/complete-bundle > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Complete Bundle All 100+ professional-grade marketing skills in one package. ## Common Parameters All skills accept: - `companyName` (string, required): Company or product name - `description` (string, required): Product or service description - Additional skill-specific parameters (all optional unless noted) --- ## Business Foundations (5 skills) ### generate_persona Create detailed buyer personas with psychographic profiles, pain points, goals, and communication preferences. ### create_positioning Generate strategic positioning framework including positioning statement, key differentiators, and competitive angles. ### create_messaging Build comprehensive messaging framework with key messages, proof points, and audience-specific variations. ### generate_value_proposition Create value proposition ladder with benefit hierarchies from features to emotional benefits. ### create_brand_voice Define brand voice guidelines with tone, vocabulary, and writing style recommendations. --- ## Creative & Copy (5 skills) ### generate_headlines Generate compelling headlines, taglines, and hooks for various marketing channels using proven formulas (PAS, AIDA, How-To, etc.). ### generate_taglines Create memorable brand taglines and slogans that capture brand essence. ### generate_offers_hooks Create irresistible offers and psychological hooks to drive conversions. ### create_marketing_copy Generate versatile marketing copy for various use cases. ### create_product_description Write compelling product descriptions that sell. --- ## Advertising (5 skills) ### create_google_ads Generate Google Ads campaigns with headlines, descriptions, and keyword suggestions. ### create_meta_ads Generate Meta (Facebook/Instagram) ad campaigns with copy, headlines, and audience targeting. ### generate_ad_copy Create versatile ad copy for multiple advertising platforms. ### create_campaign_ideas Generate 5 creative campaign concepts with angles, billboard concepts, thought-provoking statements, and AI image prompts. ### create_ab_testing_plan Create a structured A/B testing plan with hypotheses, test designs, sample sizes, and success metrics. --- ## Content & SEO (11 skills) ### plan_content_strategy Generate comprehensive content strategy with topics, calendar, and distribution plan. ### create_seo_strategy Analyze and create SEO strategy with keyword research and content recommendations. ### generate_social_posts Create engaging social media posts for various platforms. ### optimize_for_llms Analyze and optimize content for LLM and AI search visibility (ChatGPT, Perplexity, etc.). ### create_white_paper Generate white paper outlines and content for thought leadership. ### create_case_study Generate case study frameworks and content templates. ### create_programmatic_seo Design a programmatic SEO strategy with template types, data sources, URL structures, and scalable page-generation blueprint. ### generate_article_from_topic Create structured article briefs and outlines for any topic, ready to publish. ### research_facts_for_content Surface credible statistics, studies, and sources to back your content arguments. ### write_long_form_draft Generate a complete long-form article or guide with full sections, narrative flow, and SEO structure. ### hunt_viral_stories Identify trending angles, story pegs, and cultural hooks to ride for content amplification. --- ## Website & Landing Pages (3 skills) ### create_landing_page Generate high-converting landing page copy with headlines, benefits, and CTAs. ### analyze_homepage Analyze homepage messaging and provide conversion optimization recommendations. ### generate_faqs Create comprehensive FAQ content addressing common customer questions. --- ## Sales & Email (6 skills) ### create_lifecycle_emails Generate lifecycle email sequences for onboarding, nurturing, and retention. ### create_sales_script Generate sales scripts and talk tracks for different scenarios. ### create_cold_email_sequence Generate cold outbound email sequences with personalization frameworks, follow-up cadences, and reply handling. ### create_buyers_journey Map the full buyer's journey with awareness, consideration, and decision-stage content, messages, and objection handling. ### sales_roleplay Simulate realistic buyer conversations to sharpen objection handling and discovery skills. ### find_icp_leads Generate Boolean search strings, trigger events, and prospecting filters to build targeted lead lists. --- ## Video & Media (4 skills) ### create_video_script Generate video scripts for marketing, explainer, and promotional videos. ### create_press_release Generate professional press releases for company announcements. ### find_media_outlets Identify relevant media outlets and journalists for PR outreach. ### create_media_angles Generate PR angles and story pitches for media outreach. --- ## Research & Intelligence (2 skills) ### analyze_competitor Deep competitive analysis with positioning gaps, SWOT, and battlecard materials. ### create_competitive_intelligence Generate competitive intelligence briefs and positioning gaps. --- ## Strategy (8 skills) ### create_gtm_strategy Generate a full Go-to-Market strategy with ICP targeting, channel mix, launch phases, and success metrics. ### create_gtm_motion Define the optimal GTM motion (PLG, sales-led, channel-led, community-led) with rationale and execution blueprint. ### create_channel_matrix Build a prioritized channel matrix with effort/impact scoring, budget allocation, and execution playbook per channel. ### create_pipeline_playbook Generate a complete pipeline playbook with stage definitions, entry/exit criteria, conversion tactics, and velocity benchmarks. ### create_icp_sizing Size your ICP segments with TAM/SAM/SOM analysis, firmographic filters, and qualification criteria. ### create_category_positioning Define and own a market category with naming, narrative, and POV strategy to create a new competitive frame. ### create_ecosystem_messaging Build messaging for partner ecosystems, integration networks, and platform plays with co-marketing angles. ### create_feature_messaging Translate technical features into market-facing positioning with proof points and buyer relevance. --- ## Growth & CRO (5 skills) ### create_referral_program Design a viral referral program with incentive structures, K-factor optimization, and referral mechanics. ### create_pricing_strategy Develop pricing strategy with models, tiers, and competitive positioning. ### create_analytics_tracking Create a GA4/GTM tracking plan with event taxonomy, conversion funnels, and implementation guide. ### analyze_page_cro Analyze any marketing page for conversion rate optimization opportunities with prioritized recommendations. ### create_signup_flow_cro Optimize signup and onboarding flows with friction analysis, copy improvements, and activation tactics. --- ## Output Rules - ✓ No opening comments or fluff - ✓ Direct substantive content - ✓ 400+ lines of structured output (where applicable) - ✓ Auto-inference of B2B/B2C/D2C context - ✓ Credible reasoning, no hallucinated statistics ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this complete marketing bundle to my {{companyName}} project" - **run_launch_kit** (green, primary) — sendPrompt: "Use this bundle to build the {{companyName}} launch kit — positioning, landing, ads, emails" - **run_content_engine** (neutral, secondary) — sendPrompt: "Build a 90-day content engine for {{companyName}} grounded in this bundle" - **create_gtm_strategy** (neutral, secondary) — sendPrompt: "Translate this bundle into a GTM strategy for {{companyName}} with channels and KPIs" - **run_sales_enablement** (neutral, secondary) — sendPrompt: "Build the sales enablement kit for {{companyName}} on top of this bundle" --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Content Seo Source: https://maestrix.ai/skills/content-seo > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Content & SEO Professional-grade content and SEO tools for building authority, driving organic traffic, and optimizing for AI search. ## CONTEXT NEEDED - **Company/product**: What you sell and who buys it - **Content goal**: Organic growth, thought leadership, social engagement, or LLM visibility - **Target audience**: Roles, industries, or buyer personas - **Keywords or themes** *(optional)*: Priority topics, clusters, or search terms - **Existing content gaps** *(optional)*: What's missing from your current library --- ## Available Tools (6) ### 1. create_content_strategy Build comprehensive content strategies with pillar pages, topic clusters, and editorial calendars. ### 2. generate_seo_strategy Create complete SEO strategies with keyword research, on-page optimization, and link building plans. ### 3. create_social_posts Generate platform-optimized social media content for LinkedIn, Twitter/X, Instagram, and more. ### 4. generate_llm_optimization Create strategies for optimizing content visibility in AI search results (ChatGPT, Perplexity, Gemini, etc.). ### 5. create_white_paper Generate comprehensive white papers and research reports with executive summaries. ### 6. generate_case_study Create compelling customer case studies with problem-solution-results format. --- ## Output Rules - ✓ No opening comments or fluff - ✓ Direct substantive content - ✓ SEO best practices embedded throughout - ✓ Platform-specific formatting for social content - ✓ Actionable recommendations with clear next steps --- ## Example Prompts 1. **Content Strategy**: "Create a 90-day content strategy for our HR tech startup" 2. **SEO Strategy**: "Generate an SEO plan targeting 'employee engagement software'" 3. **Social Posts**: "Create a week of LinkedIn posts announcing our Series A funding" 4. **LLM Optimization**: "How do we appear in ChatGPT responses for 'best CRM software'?" 5. **White Paper**: "Create a white paper on 'The Future of Remote Work Productivity'" 6. **Case Study**: "Write a case study about how Acme Corp increased productivity 40%" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this content SEO plan to my {{companyName}} project" - **create_content_plan** (green, primary) — sendPrompt: "Build a 90-day content calendar for {{companyName}} from this SEO plan" - **generate_article_from_topic** (neutral, secondary) — sendPrompt: "Write the first cluster article for {{companyName}} from this SEO plan" - **create_seo_strategy** (neutral, secondary) — sendPrompt: "Go deeper on the {{companyName}} SEO strategy — backlinks, topical authority, technical SEO" - **create_llm_optimization** (neutral, secondary) — sendPrompt: "Optimize {{companyName}} content for LLM retrieval and citation" --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Create Ab Testing Plan Source: https://maestrix.ai/skills/create-ab-testing-plan > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive A/B testing strategy with prioritized experiments and measurement frameworks. ## CONTEXT NEEDED - Company name and description - Page/feature to optimize - Current conversion rate (if known) - Primary KPI to improve - Traffic volume (optional) ## A/B TESTING FRAMEWORK ### 1. Testing Opportunity Audit | Page/Element | Current State | Hypothesis Category | Impact Potential | Priority | |--------------|---------------|---------------------|------------------|----------| Identify 10-15 testing opportunities. ### 2. Hypothesis Framework For each test (10 recommended tests): **Test Name**: [Descriptive name] **Hypothesis Statement** If we [change], then [metric] will [improve] because [reason based on user psychology/data]. **Variable Type**: - Copy | Design | Layout | Functionality | Offer **Test Element Details** - Control (A): Current state - Variant (B): Proposed change - Variant (C): Alternative (if applicable) **Primary Metric**: [Conversion to track] **Secondary Metrics**: [Supporting metrics] **Guardrail Metrics**: [Metrics that shouldn't decrease] **Audience Segment**: Who sees this test **Traffic Allocation**: % per variant **Minimum Sample Size**: Calculated requirement **Estimated Duration**: Based on traffic ### 3. Prioritization Matrix (PIE Framework) | Test | Potential (1-10) | Importance (1-10) | Ease (1-10) | PIE Score | Rank | |------|------------------|-------------------|-------------|-----------|------| ### 4. Testing Roadmap **Month 1**: [Tests 1-3] **Month 2**: [Tests 4-6] **Month 3**: [Tests 7-9] ### 5. Statistical Requirements | Test | Min Sample/Variant | MDE | Confidence Level | Duration Estimate | |------|-------------------|-----|------------------|-------------------| **Definitions**: - MDE: Minimum Detectable Effect - Recommended confidence: 95% - Power: 80% ### 6. Test Documentation Template For each live test, document: - Start date - Screenshot of variants - Targeting rules - Success criteria - Decision date - Results summary - Next steps ### 7. Analysis Framework **Pre-Test Checklist** - [ ] AA test passed - [ ] Tracking verified - [ ] Sample size calculated - [ ] Segments defined **Post-Test Analysis** - Statistical significance achieved? - Confidence interval - Effect size (lift %) - Segment breakdowns - Secondary metric impacts - Revenue impact projection ### 8. Decision Framework | Outcome | Action | |---------|--------| | Clear winner (95%+ confidence) | Implement | | Directional winner (90-95%) | Extend test or replicate | | No difference | Inconclusive - test bolder variant | | Loser | Learn and iterate | ### 9. Iteration Strategy Based on each test outcome: - What to test next - How to compound wins - When to stop testing an area ### 10. Reporting Template Weekly testing dashboard: - Active tests - Tests completed - Win rate - Cumulative lift - Revenue impact ## QUALITY RULES 1. NO opening comments 2. Statistically valid sample sizes 3. Clear hypothesis with rationale 4. Actionable follow-up recommendations ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C **Layout spec:** Header row (one card per competitor with score badge) → feature matrix (color-coded: --success win, --destructive lose, --muted neutral) → battle card block (positioning, weakness, counter) → bottom highlights (--warning gap, --success opportunity). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create the landing page variation for {{companyName}} — {{description}}, to test first' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, to drive test traffic' - Label: 'Analytics tracking ↗' sendPrompt: 'Create an analytics tracking plan for {{companyName}} — {{description}}, to measure test results' - Label: 'CRO audit ↗' sendPrompt: 'Run a CRO audit for {{companyName}} — {{description}}, to identify more optimization opportunities' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Ad Copy Source: https://maestrix.ai/skills/create-ad-copy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create high-converting ad copy adaptable across multiple advertising platforms. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Campaign objective (awareness, consideration, conversion) - Platform focus (optional) ## AD COPY FRAMEWORK ### 1. Campaign Brief Summary - Objective - Target audience - Key message - Desired action - Constraints/requirements ### 2. Core Message Variations **Problem-Agitation Headlines** (5) **Solution-Benefit Headlines** (5) **Curiosity-Gap Headlines** (5) **Social Proof Headlines** (5) **Urgency Headlines** (5) ### 3. Platform-Specific Copy **LinkedIn Ads** - Sponsored Content (intro text, 150 chars) - Headlines (5 options, 70 chars each) - Descriptions (5 options, 100 chars each) - Message Ads intro (300 words) **Twitter/X Ads** - Promoted tweets (5 options, 280 chars) - Carousel headlines (5 cards) - Website card copy **Meta Ads (Facebook/Instagram)** - Primary text (5 options, 125 chars optimal) - Headlines (5 options, 40 chars) - Descriptions (5 options, 30 chars) - Story ad copy **TikTok Ads** - Hook lines (5 options, first 3 seconds) - CTA overlays - Caption copy **Google Display** - Short headlines (5, 30 chars) - Long headlines (5, 90 chars) - Descriptions (5, 90 chars) ### 4. Funnel Stage Variations **TOFU (Awareness)** - Educational angle - Problem-aware messaging - Broad appeal copy **MOFU (Consideration)** - Comparison messaging - Feature highlights - Social proof emphasis **BOFU (Conversion)** - Offer-focused - Urgency/scarcity - Direct CTA ### 5. A/B Testing Matrix | Test Variable | Version A | Version B | Hypothesis | |---------------|-----------|-----------|------------| Provide 5 testable hypotheses with copy variants. ### 6. CTA Variations - Direct CTAs (5) - Soft CTAs (5) - Value-add CTAs (5) ### 7. Visual Direction For each copy set, suggest: - Image/video concept - Text overlay recommendations - Color/mood direction ### 8. Audience Segment Adaptations Copy variations for: - Segment A (description + adapted copy) - Segment B (description + adapted copy) - Segment C (description + adapted copy) ### 9. Seasonal/Timely Hooks - Evergreen versions - Q1-Q4 adaptations - Event-based variations ### 10. Compliance Notes - Claims to verify - Disclaimers needed - Platform-specific restrictions ## QUALITY RULES 1. NO opening comments 2. Character limits strictly enforced 3. Multiple variations per element 4. Clear funnel stage labeling ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a destination for these ads' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, complementing this ad copy' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to test ad variations' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, reinforcing the ad messaging' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Analytics Tracking Source: https://maestrix.ai/skills/create-analytics-tracking > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a complete analytics tracking plan with event schemas, implementation specifications, and measurement frameworks. ## CONTEXT NEEDED - Company name and description - Primary conversion goals - Key user journeys to track - Current analytics tools (GA4, Mixpanel, Amplitude, etc.) - Technical implementation capacity ## ANALYTICS TRACKING FRAMEWORK ### 1. Measurement Strategy **Business Objectives → Metrics Mapping** | Business Objective | Key Question | Primary Metric | Secondary Metrics | |--------------------|--------------|----------------|-------------------| | | | | | **Conversion Funnel Definition** ``` [Awareness] → [Interest] → [Consideration] → [Intent] → [Purchase] → [Loyalty] ↓ ↓ ↓ ↓ ↓ ↓ [Metric] [Metric] [Metric] [Metric] [Metric] [Metric] ``` ### 2. Event Taxonomy **Naming Convention** ``` Format: [object]_[action] Examples: - page_view - button_click - form_submit - product_view - cart_add - checkout_start - purchase_complete ``` **Event Categories** | Category | Description | Examples | |----------|-------------|----------| | Pageview | Page loads | page_view | | Engagement | User interactions | scroll, click, video_play | | Conversion | Goal completions | signup, purchase, lead_submit | | E-commerce | Transaction events | add_to_cart, checkout, refund | | Error | Issues encountered | form_error, 404_error | ### 3. Core Event Schema **Page View Events** ```javascript // GA4 Enhanced Measurement (automatic) // Custom page_view for SPAs gtag('event', 'page_view', { page_title: 'Product Page - [Name]', page_location: window.location.href, page_path: '/products/[slug]', user_id: '[hashed_user_id]', // if logged in page_type: 'product', // custom dimension content_group: '[category]' }); ``` **User Journey Events** | Event Name | Trigger | Required Parameters | Optional Parameters | |------------|---------|---------------------|---------------------| | session_start | First pageview | | traffic_source, landing_page | | user_signup | Registration complete | method, user_type | referral_source | | login | User logs in | method | days_since_last | | onboarding_start | Begins onboarding | | | | onboarding_complete | Finishes onboarding | completion_time | steps_completed | | feature_use | Uses key feature | feature_name | usage_count | | upgrade_start | Begins upgrade | current_plan | | | purchase | Transaction complete | value, currency | items, coupon | **Engagement Events** | Event Name | Trigger | Parameters | |------------|---------|------------| | scroll_depth | 25%, 50%, 75%, 90% | percent, page | | video_start | Video begins | video_title, video_duration | | video_progress | 25%, 50%, 75% | video_title, percent | | video_complete | Video ends | video_title, watch_time | | file_download | Download click | file_name, file_type | | outbound_click | External link | link_url, link_text | | search | Site search | search_term, results_count | | cta_click | CTA button click | cta_text, cta_location | ### 4. E-commerce Event Schema (if applicable) **Product Interaction Events** ```javascript // View item gtag('event', 'view_item', { currency: 'USD', value: 29.99, items: [{ item_id: 'SKU_123', item_name: 'Product Name', item_category: 'Category', item_variant: 'Variant', price: 29.99, quantity: 1 }] }); // Add to cart gtag('event', 'add_to_cart', { currency: 'USD', value: 29.99, items: [/* same structure */] }); // Begin checkout gtag('event', 'begin_checkout', { currency: 'USD', value: 59.98, coupon: 'SUMMER20', items: [/* items array */] }); // Purchase gtag('event', 'purchase', { transaction_id: 'T_12345', value: 59.98, tax: 4.80, shipping: 5.00, currency: 'USD', coupon: 'SUMMER20', items: [/* items array */] }); ``` ### 5. Custom Dimensions & Metrics **User-Scoped Dimensions** | Dimension | Description | Values | Use Case | |-----------|-------------|--------|----------| | user_type | Account type | free, trial, paid | Segment by plan | | signup_date | Registration date | YYYY-MM-DD | Cohort analysis | | industry | User's industry | [list] | B2B segmentation | | company_size | Company employees | 1-10, 11-50, 51-200, 200+ | Enterprise analysis | **Event-Scoped Dimensions** | Dimension | Description | Values | Use Case | |-----------|-------------|--------|----------| | content_type | Type of content | blog, video, guide | Content performance | | feature_name | Feature used | [list] | Feature adoption | | experiment_id | A/B test ID | [test_name]_[variant] | Experiment analysis | | error_type | Error category | form, api, 404 | Error monitoring | **Custom Metrics** | Metric | Description | Calculation | |--------|-------------|-------------| | engagement_score | User engagement level | Composite of actions | | feature_adoption | Features activated | Count of first-uses | | time_to_value | Seconds to first success | timestamp delta | ### 6. UTM & Campaign Tracking **UTM Parameter Standards** | Parameter | Usage | Examples | |-----------|-------|----------| | utm_source | Traffic source | google, facebook, newsletter | | utm_medium | Marketing medium | cpc, email, social, organic | | utm_campaign | Campaign name | spring_sale_2024, product_launch | | utm_content | A/B test or creative | hero_v1, sidebar_banner | | utm_term | Paid keywords | marketing_automation | **URL Builder Template** ``` Base URL: https://example.com/landing-page Full URL: https://example.com/landing-page?utm_source=facebook&utm_medium=cpc&utm_campaign=spring_2024&utm_content=video_ad_1 ``` ### 7. GTM Implementation Guide **Container Structure** ``` GTM Container ├── Tags │ ├── GA4 Configuration │ ├── GA4 Event - Page View │ ├── GA4 Event - Button Clicks │ ├── GA4 Event - Form Submissions │ ├── GA4 Event - Scroll Depth │ ├── GA4 Event - Video Tracking │ ├── Conversion Pixels (Meta, LinkedIn, etc.) │ └── Custom HTML (if needed) ├── Triggers │ ├── All Pages │ ├── Button Clicks - CTA │ ├── Form Submission - Contact │ ├── Scroll Depth Thresholds │ ├── Video Milestones │ └── Custom Events └── Variables ├── GA4 Measurement ID ├── User ID ├── Page Type ├── Click Text ├── Click URL └── Data Layer Variables ``` **Data Layer Implementation** ```javascript // Push to data layer on key events window.dataLayer = window.dataLayer || []; // User identification dataLayer.push({ 'event': 'user_identified', 'user_id': 'hashed_id_123', 'user_type': 'paid', 'signup_date': '2024-01-15' }); // Custom event dataLayer.push({ 'event': 'cta_click', 'cta_text': 'Start Free Trial', 'cta_location': 'hero', 'page_type': 'homepage' }); ``` ### 8. Dashboard & Reporting **Executive Dashboard KPIs** | Metric | Definition | Target | Visualization | |--------|------------|--------|---------------| | Conversion Rate | Conversions / Sessions | % | Trend line | | Revenue | Total transaction value | $ | Bar chart | | CAC | Marketing spend / Customers | $ | Gauge | | LTV | Avg customer lifetime value | $ | Comparison | **Marketing Performance Dashboard** | Report | Dimensions | Metrics | Frequency | |--------|------------|---------|-----------| | Channel Performance | Source/Medium | Sessions, Conversions, Revenue | Weekly | | Campaign Analysis | Campaign | Clicks, Cost, ROAS | Weekly | | Content Performance | Page, Content Type | Views, Engagement, Conversions | Monthly | | Funnel Analysis | Step | Users, Drop-off rate | Daily | **Product Analytics Dashboard** | Report | Focus | Key Metrics | |--------|-------|-------------| | Activation | Onboarding | Completion rate, Time to activate | | Engagement | Feature use | DAU/MAU, Feature adoption | | Retention | Cohorts | D1, D7, D30 retention | | Conversion | Upgrade | Trial-to-paid, Expansion revenue | ### 9. Implementation Checklist **Phase 1: Foundation (Week 1)** - [ ] GA4 property created and configured - [ ] GTM container installed on all pages - [ ] Enhanced measurement enabled - [ ] Cross-domain tracking configured (if needed) - [ ] User ID tracking implemented - [ ] Data retention settings configured **Phase 2: Core Events (Week 2)** - [ ] Conversion events defined - [ ] E-commerce tracking implemented - [ ] Form submission tracking - [ ] CTA click tracking - [ ] Custom dimensions configured **Phase 3: Advanced Tracking (Week 3)** - [ ] Scroll depth tracking - [ ] Video engagement tracking - [ ] File download tracking - [ ] Error tracking - [ ] Search tracking **Phase 4: Reporting (Week 4)** - [ ] Looker Studio dashboards created - [ ] Custom reports configured - [ ] Automated alerts set up - [ ] Documentation completed ### 10. Privacy & Compliance **Consent Management** - Cookie consent banner implemented - Analytics only fires after consent - Consent mode v2 configured - Data retention policies set **Data Collection Best Practices** - No PII in custom dimensions - User IDs are hashed - IP anonymization enabled - Data deletion processes documented ## QUALITY RULES 1. NO opening comments - start with measurement strategy 2. Use standard GA4 event naming conventions 3. Include actual code snippets for implementation 4. Provide complete data layer specifications 5. Include privacy/compliance considerations ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, with tracking built in from day one' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, with proper conversion tracking' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, instrumented with this tracking plan' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, using this analytics framework' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Brand Voice Source: https://maestrix.ai/skills/create-brand-voice > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive brand voice guide that defines how the brand communicates across all channels. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Existing brand values (optional) - Competitor examples (optional) ## BRAND VOICE FRAMEWORK ### 1. Voice Personality Traits Define 4-5 core personality traits with: - Trait name - Definition (what it means for this brand) - In practice (how it manifests in communication) - Spectrum positioning (e.g., Formal ←→ Casual: 7/10) ### 2. Tone Dimensions Rate each dimension 1-10: - Formal vs. Casual - Serious vs. Playful - Respectful vs. Irreverent - Enthusiastic vs. Matter-of-fact - Technical vs. Accessible ### 3. Voice Do's and Don'ts | Do | Don't | |----|----| | [Specific behavior] | [Opposite behavior] | Provide 10+ pairs covering vocabulary, sentence structure, humor, jargon, and emotional expression. ### 4. Vocabulary Guidelines **Power Words**: 15-20 words that embody the brand **Forbidden Words**: Words to avoid and why **Industry Jargon**: When to use vs. simplify **Branded Terms**: Proprietary language and usage rules ### 5. Channel-Specific Adaptations For each channel (Website, Email, Social, Support, Sales): - Tone adjustment - Length guidelines - Formality level - Example snippet ### 6. Audience Segment Variations How voice shifts for: - Prospects vs. Customers - Technical vs. Non-technical - Enterprise vs. SMB - Different personas ### 7. Sample Copy Transformations Show 5 examples: - Generic version - On-brand version - Why it works ### 8. Voice Checklist 10-item checklist for content review ensuring brand voice consistency. ## QUALITY RULES 1. NO opening comments 2. Specific, actionable guidelines (not vague) 3. Real examples for every concept 4. Measurable criteria where possible ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, using the brand voice we just defined' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, following this brand voice' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, in the brand voice we just defined' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, written in this brand voice' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Buyers Journey Source: https://maestrix.ai/skills/create-buyers-journey > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive buyer's journey map with stages, touchpoints, emotions, and content needs. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience/persona - Sales cycle length (optional) - Key competitors (optional) ## BUYER'S JOURNEY FRAMEWORK ### 1. Journey Stage Overview | Stage | Duration | Key Question | Emotional State | Success Metric | |-------|----------|--------------|-----------------|----------------| | Unaware | | | | | | Problem Aware | | | | | | Solution Aware | | | | | | Product Aware | | | | | | Most Aware | | | | | ### 2. Detailed Stage Analysis For each of the 5 stages: **Stage Name** - **Trigger**: What moves them into this stage - **Mindset**: Internal dialogue and concerns - **Questions Asked**: 5-7 specific questions - **Information Sources**: Where they seek answers - **Influencers**: Who affects their thinking - **Barriers**: What prevents progression - **Exit Trigger**: What moves them to next stage ### 3. Touchpoint Mapping | Touchpoint | Stage | Channel | Purpose | Content Type | Conversion Goal | |------------|-------|---------|---------|--------------|-----------------| List 15-20 touchpoints across the journey. ### 4. Content Needs Matrix | Stage | Content Type | Topic Focus | Format | Distribution | |-------|--------------|-------------|--------|--------------| Map specific content pieces to each stage (20+ items). ### 5. Objection Timeline When objections arise and how to address: - Stage 1 objections and responses - Stage 2 objections and responses - Stage 3 objections and responses - Stage 4 objections and responses - Stage 5 objections and responses ### 6. Conversion Triggers Identify 10+ psychological triggers that drive stage progression: - Trigger name - Stage transition it enables - How to activate it - Content/tactic example ### 7. Friction Points & Solutions | Friction Point | Stage | Impact | Solution | Priority | |----------------|-------|--------|----------|----------| ### 8. Measurement Framework KPIs for each stage: - Leading indicators - Lagging indicators - Conversion rates - Time-in-stage benchmarks ## QUALITY RULES 1. NO opening comments 2. Specific to the company/product context 3. Actionable content recommendations 4. Measurable outcomes for each stage ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Sales script ↗' - sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, mapped to this buyer's journey' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, with content for each journey stage' - Label: 'Lifecycle emails ↗' sendPrompt: 'Create lifecycle emails for {{companyName}} — {{description}}, nurturing buyers through each stage' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, for early-stage prospects' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Campaign Ideas Source: https://maestrix.ai/skills/create-campaign-ideas > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create innovative marketing campaign concepts with full execution frameworks. ## CONTEXT NEEDED - Company name and description - Campaign objective - Target audience - Budget range (optional) - Timeline (optional) - Brand guidelines/constraints (optional) ## CAMPAIGN IDEATION FRAMEWORK ### 1. Campaign Concepts (5 Complete Ideas) For each concept: **Campaign Name** Creative, memorable title **The Big Idea** One-sentence concept summary **Insight** Customer/market insight that inspired it **Theme** Visual and messaging theme **Tagline** Campaign tagline (3-5 words) **Emotional Hook** Primary emotion targeted **Key Message** What audience should remember **Mechanics** How the campaign works **Channels** Primary and supporting channels **Timeline** Suggested campaign duration **Hero Content** Main creative asset(s) **Supporting Assets** Secondary content pieces **Interactive Element** Engagement/participation component **Measurement** Primary success metrics ### 2. Creative Concepts Comparison | Concept | Novelty | Feasibility | Brand Fit | Viral Potential | Recommended? | |---------|---------|-------------|-----------|-----------------|--------------| ### 3. Selected Concept Deep Dive (Expand on #1 recommended concept) **Creative Brief** - Background - Objective - Target audience - Key message - Tone and manner - Mandatories - Deliverables **Content Calendar** | Week | Channel | Content | Purpose | CTA | |------|---------|---------|---------|-----| **Influencer/Partner Integration** - Partner types - Collaboration concepts - Outreach approach **User-Generated Content Strategy** - UGC prompt - Hashtag - Incentive structure - Curation plan **PR Angle** - Newsworthy element - Story pitch - Target publications ### 4. Channel Execution Details For each primary channel: - Creative specifications - Copy direction - Targeting approach - Budget allocation % - Success metrics ### 5. Launch Sequence **Pre-Launch Tease** (Week -1) **Launch Day** (Day 1) **Momentum Phase** (Days 2-7) **Sustain Phase** (Weeks 2-4) **Close/Transition** ### 6. Risk Assessment | Risk | Mitigation | |------|------------| ### 7. Budget Allocation Recommendation | Category | % | Activities | |----------|---|------------| ### 8. Scaling Scenarios **Minimum Viable Campaign** **Full Execution** **Expanded/Viral Response** ## QUALITY RULES 1. NO opening comments 2. Bold, differentiated ideas 3. Actionable execution details 4. Clear measurement framework ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building on the campaign ideas we just generated' Secondary actions (neutral styling): - Label: 'Meta Ads ↗' sendPrompt: 'Create Meta Ads for {{companyName}} — {{description}}, based on the best campaign concept' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, aligned with these campaign ideas' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, to support these campaign concepts' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Case Study Source: https://maestrix.ai/skills/create-case-study > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create compelling customer case studies that demonstrate value and drive conversions. ## CONTEXT NEEDED - Company name and description - Customer name and industry - Solution/product used - Results achieved (metrics) - Customer quotes (optional) ## CASE STUDY FRAMEWORK ### 1. Case Study Brief **Title Options** (5) - Result-focused: "[Company] Achieves [Result] with [Product]" - Challenge-focused: "How [Company] Solved [Problem]" - Transformation: "[Company]: From [Before] to [After]" **Subtitle** Specific hook or key metric highlight **At-a-Glance Box** - Company: [Name] - Industry: [Industry] - Size: [Employees/Revenue] - Challenge: [One line] - Solution: [Product/Service] - Results: [Key metrics] ### 2. Story Structure **1. The Customer (200 words)** - Company background - Industry context - Business model - Why they're a good reference **2. The Challenge (400 words)** - Situation before - Specific pain points - Failed previous attempts - Stakes/urgency - Quote: Challenge in customer's words **3. The Solution (500 words)** - Why they chose your solution - Implementation process - Key features/capabilities used - Integration with existing systems - Quote: Decision-making insight **4. The Implementation (300 words)** - Timeline - Team involved - Training/onboarding - Any challenges overcome - Quote: Implementation experience **5. The Results (500 words)** - Primary metric improvements - Secondary benefits - Unexpected wins - Before/after comparison - Quote: Results in customer's words **6. The Future (200 words)** - Expansion plans - Additional use cases - Long-term partnership - Quote: Forward-looking statement ### 3. Metrics Showcase | Metric | Before | After | Improvement | |--------|--------|-------|-------------| | | | | | Feature 3-5 quantified results. ### 4. Quote Library Extract/create quotes for: - Challenge acknowledgment - Solution selection - Implementation experience - Results celebration - Recommendation to others ### 5. Visual Storyboard | Section | Visual | Purpose | |---------|--------|---------| | Company | Logo, office photo | Establish credibility | | Challenge | Before state visual | Show pain | | Solution | Product screenshot | Show solution | | Results | Chart/graph | Prove value | | Team | Customer headshot | Add human element | ### 6. Format Variations **Long-Form PDF** (2000+ words) - Full narrative - Design-heavy - Gated asset **Web Page Version** (1200 words) - Scannable - Above-fold results - Embedded video option **One-Pager** (400 words) - Sales enablement - Quick reference - Key stats focus **Video Script** (3-5 minutes) - Customer interview Qs - B-roll shot list - Key soundbites **Social Snippets** - LinkedIn post - Twitter thread - Quote graphics ### 7. SEO Optimization - Target keywords - Meta title - Meta description - H1/H2 structure - Internal links - Schema markup ### 8. Distribution Plan **Sales Enablement** - When to use - How to position - Follow-up sequence **Marketing Channels** - Website placement - Email campaigns - Social distribution - Paid promotion ### 9. Customer Approval Checklist - [ ] Accuracy review - [ ] Quote approval - [ ] Metric verification - [ ] Logo/photo permission - [ ] Legal sign-off ## QUALITY RULES 1. NO opening comments 2. Specific metrics, not vague improvements 3. Customer voice prominent (quotes) 4. Clear before/after narrative ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, featuring this case study as social proof' Secondary actions (neutral styling): - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, incorporating the case study results' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this case study' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, leading with these case study results' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Category Positioning Source: https://maestrix.ai/skills/create-category-positioning > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a category positioning strategy that establishes market leadership through category design. ## CONTEXT NEEDED - Company name and description - Products or services offered - Current market category - Target audience - Key competitors (optional) ## CATEGORY POSITIONING FRAMEWORK ### 1. Current Category Analysis **Existing Category Landscape** - Category name and definition - Category size and growth rate - Key players and market share - Category conventions and expectations - Buyer decision criteria - Category limitations ### 2. Category Opportunity Assessment | Opportunity Type | Description | Viability Score (1-10) | |------------------|-------------|------------------------| | Category Creation | Define entirely new space | | | Category Redesign | Reframe existing category | | | Subcategory Carve-out | Own a segment | | | Category Combination | Merge two categories | | Recommend best approach with rationale. ### 3. New Category Design **Category Name Options** (5 alternatives): - Name - Rationale - Memorability score - Search viability **Category Definition** - One-sentence definition - Problem it solves - Who it's for - What it replaces ### 4. Category Point of View (POV) **The Manifesto** - What's broken in the current approach - Why now is the time for change - What the new category enables - Vision of the future state ### 5. Category Differentiation Matrix | Attribute | Old Category | New Category | Why It Matters | |-----------|--------------|--------------|----------------| List 8-10 key differentiators. ### 6. Lightning Strike Strategy **Phase 1: Category Introduction** (Month 1-3) - Key activities - Content themes - Target audiences **Phase 2: Category Evangelism** (Month 4-6) - Thought leadership plan - Analyst engagement - Community building **Phase 3: Category Dominance** (Month 7-12) - Market education - Ecosystem development - Competitive response plan ### 7. Category Ecosystem - Complementary products/services - Potential partners - Analyst firms to educate - Media outlets to pitch - Events to sponsor/create ### 8. Messaging Architecture **Category-level messaging** - Headline - Subhead - Proof points **Company-level messaging** - Why we're the category leader - Credibility builders ### 9. Competitive Response Playbook How to respond when competitors: - Ignore the category - Attack the category - Try to co-opt the category - Create counter-category ## QUALITY RULES 1. NO opening comments 2. Bold, differentiated positioning 3. Actionable go-to-market tactics 4. Realistic timeline and milestones ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, owning this new category' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, establishing category thought leadership' - Label: 'Press release ↗' sendPrompt: 'Write a press release for {{companyName}} — {{description}}, announcing this category' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, defining this category' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Channel Matrix Source: https://maestrix.ai/skills/create-channel-matrix > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive channel strategy with prioritization, budget allocation, and execution playbooks. ## CONTEXT NEEDED - Company name and description - Target audience / ICP - Monthly/quarterly marketing budget - Current channels in use - Sales cycle length - Average deal size / LTV ## CHANNEL STRATEGY FRAMEWORK ### 1. Channel Universe Mapping Score each channel (1-10) on three dimensions: **Paid Channels** | Channel | ICP Fit | Scalability | CAC Efficiency | Total | Priority | |---------|---------|-------------|----------------|-------|----------| | Google Search Ads | | | | /30 | | | Google Display | | | | /30 | | | Meta (FB/IG) | | | | /30 | | | LinkedIn Ads | | | | /30 | | | YouTube Ads | | | | /30 | | | TikTok Ads | | | | /30 | | | Reddit Ads | | | | /30 | | | Podcast Ads | | | | /30 | | | Programmatic Display | | | | /30 | | | Influencer Paid | | | | /30 | | **Organic Channels** | Channel | ICP Fit | Scalability | CAC Efficiency | Total | Priority | |---------|---------|-------------|----------------|-------|----------| | SEO / Content | | | | /30 | | | LinkedIn Organic | | | | /30 | | | Twitter/X Organic | | | | /30 | | | YouTube Organic | | | | /30 | | | Community Building | | | | /30 | | | Newsletter | | | | /30 | | | Podcast (Owned) | | | | /30 | | **Outbound Channels** | Channel | ICP Fit | Scalability | CAC Efficiency | Total | Priority | |---------|---------|-------------|----------------|-------|----------| | Cold Email | | | | /30 | | | Cold Calling | | | | /30 | | | LinkedIn Outreach | | | | /30 | | | Direct Mail | | | | /30 | | | Account-Based (ABM) | | | | /30 | | **Inbound Channels** | Channel | ICP Fit | Scalability | CAC Efficiency | Total | Priority | |---------|---------|-------------|----------------|-------|----------| | Webinars | | | | /30 | | | Events (Virtual) | | | | /30 | | | Events (In-Person) | | | | /30 | | | Referral Program | | | | /30 | | | Partnerships | | | | /30 | | | Product Virality | | | | /30 | | ### 2. Priority Tier Assignment **Tier 1: Core Channels** (60-70% of budget) Top 2-3 channels for consistent investment. | Channel | Monthly Budget | Expected CAC | Target Leads/Mo | Owner | |---------|----------------|--------------|-----------------|-------| **Tier 2: Growth Channels** (20-30% of budget) Scaling channels showing early promise. | Channel | Monthly Budget | Expected CAC | Target Leads/Mo | Owner | |---------|----------------|--------------|-----------------|-------| **Tier 3: Experimental Channels** (10% of budget) Testing new opportunities. | Channel | Monthly Budget | Test Hypothesis | Success Criteria | Timeline | |---------|----------------|-----------------|------------------|----------| ### 3. Channel Mix by Buyer Journey Stage | Stage | Primary Channels | Content/Offer | Conversion Goal | |-------|------------------|---------------|-----------------| | Awareness | | | | | Consideration | | | | | Decision | | | | | Retention | | | | | Advocacy | | | | ### 4. Budget Allocation Model **Total Monthly Budget:** $[X] | Category | % | Amount | Rationale | |----------|---|--------|-----------| | Paid Acquisition | | $ | | | Content/SEO | | $ | | | Outbound | | $ | | | Events | | $ | | | Partnerships | | $ | | | Tools/Tech | | $ | | | Reserve/Testing | | $ | | ### 5. Channel Playbooks #### Playbook: [Top Priority Channel] **Setup Requirements** - Tools needed: - Team capacity: - Content requirements: - Tracking setup: **Targeting Strategy** - Audience definition: - Exclusions: - Lookalike/Similar audiences: **Creative Strategy** - Ad formats: - Messaging themes: - Testing cadence: **Optimization Cadence** - Daily: Check spend, pause underperformers - Weekly: Adjust bids, refresh creative - Monthly: Strategy review, budget reallocation **Success Metrics** | Metric | Target | Measurement | |--------|--------|-------------| [Repeat for each Tier 1 channel] ### 6. Channel Sequencing Plan **Month 1-2: Foundation** - Focus channels: - Key activities: - Investment level: **Month 3-4: Optimization** - Focus channels: - Key activities: - Investment level: **Month 5-6: Expansion** - Focus channels: - Key activities: - Investment level: ### 7. Attribution Model **Recommended Model:** [First-touch / Last-touch / Multi-touch / Data-driven] **Attribution Windows** | Channel Type | Click Window | View Window | |--------------|--------------|-------------| | Paid Search | 30 days | N/A | | Paid Social | 28 days | 1 day | | Display | 30 days | 7 days | | Email | 7 days | N/A | ### 8. Channel Synergies | Channel A | Channel B | Synergy Effect | How to Leverage | |-----------|-----------|----------------|-----------------| ### 9. Kill Criteria When to stop investing in a channel: | Signal | Threshold | Action | |--------|-----------|--------| | CAC exceeds LTV | 1.5x target | Pause, diagnose | | Declining efficiency | 3 consecutive months | Reduce budget 50% | | Quality issues | <20% SQL rate | Pause, requalify | | Market saturation | Diminishing returns | Shift to new segment | ### 10. Quarterly Review Template **Channel Performance Summary** | Channel | Spend | Leads | CAC | Pipeline | ROI | |---------|-------|-------|-----|----------|-----| **Key Learnings** - What worked: - What didn't: - Surprises: **Next Quarter Adjustments** - Channels to scale: - Channels to cut: - New channels to test: ## QUALITY RULES 1. NO opening comments 2. Score all channels before prioritizing 3. Budget must sum to 100% 4. Include kill criteria for discipline 5. Attribution model must match sales cycle ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, focused on the priority channels from this matrix' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, for the top paid channel' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, for the outbound channel' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, for the organic channels' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Cold Email Source: https://maestrix.ai/skills/create-cold-email > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create high-performing cold email sequences with personalization and follow-up strategies. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience/ICP - Value proposition - Social proof elements (optional) ## COLD EMAIL FRAMEWORK ### 1. Campaign Strategy **Objective**: [Meeting, demo, response, etc.] **Target Persona**: [Role, industry, company size] **Sequence Length**: [Number of emails] **Cadence**: [Timing between emails] ### 2. Personalization Framework **Research Triggers** - Company news - Role changes - Funding announcements - Product launches - Content engagement - Mutual connections - Shared experiences **Personalization Variables** | Variable | Source | Example | |----------|--------|---------| | {{company_news}} | News search | "Congrats on the Series B" | | {{pain_point}} | Industry research | "Managing remote teams" | | {{mutual}} | LinkedIn | "We both know Sarah" | | {{content}} | Their posts | "Your article on X" | ### 3. Email Sequence (5-7 Emails) **Email 1: Initial Outreach** - Subject lines (5 options, <50 chars) - Opening line (personalized hook) - Problem statement - Solution tease - Soft CTA - Signature **Email 2: Value-Add Follow-Up** (Day 3) - Subject lines (3 options) - Reference to Email 1 - Educational content/insight - Resource offer - Soft CTA **Email 3: Social Proof** (Day 7) - Subject lines (3 options) - Customer story snippet - Results/metrics - Pattern interrupt - CTA **Email 4: Different Angle** (Day 10) - Subject lines (3 options) - New pain point - Different value prop - CTA **Email 5: Breakup** (Day 14) - Subject lines (3 options) - Acknowledge no response - Final value statement - Easy out - Future door-opener ### 4. Subject Line Bank **Curiosity-Based** (10) **Personalized** (10) **Question-Based** (10) **Direct** (10) **Pattern Interrupt** (10) ### 5. Opening Lines **Personalized Openers** (10) Reference their world **Pattern Interrupts** (5) Unexpected starts **Observation Openers** (5) Notice something relevant **Mutual Connection** (5) Shared relationships ### 6. CTA Variations **Soft CTAs** - "Worth a conversation?" - "Open to learning more?" - "Would it make sense to chat?" **Specific CTAs** - "Do you have 15 min this Thursday?" - "Can I send a 2-min video?" - "Would a case study help?" **Alternative CTAs** - "Who should I talk to about this?" - "Is this on your radar for Q2?" ### 7. Objection Preemption Address common objections within emails: | Objection | How to Preempt | |-----------|----------------| | Not interested | Lead with value, not pitch | | No budget | ROI focus | | No time | Efficiency angle | | Happy with current | Innovation angle | ### 8. A/B Testing Plan | Test | Version A | Version B | Send % | |------|-----------|-----------|--------| | Subject | | | | | Opening | | | | | CTA | | | | | Length | | | | | Send time | | | | ### 9. Deliverability Best Practices - Word count targets - Link limits - Image guidance - Spam trigger words to avoid - Domain warming schedule ### 10. Performance Benchmarks | Metric | Target | Action if Below | |--------|--------|-----------------| | Open rate | 40%+ | Fix subject lines | | Reply rate | 5%+ | Fix messaging/targeting | | Meeting rate | 1%+ | Fix CTA/offer | ## QUALITY RULES 1. NO opening comments 2. Conversational, not salesy 3. Short paragraphs (2-3 sentences max) 4. One clear CTA per email ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Sales script ↗' - sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, following up on the cold email sequence we just wrote' Secondary actions (neutral styling): - Label: 'Pipeline playbook ↗' sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, incorporating this outbound email strategy' - Label: 'LinkedIn Ads ↗' sendPrompt: 'Create LinkedIn Ads for {{companyName}} — {{description}}, targeting the same audience as these cold emails' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a destination for this cold email campaign' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Competitive Intelligence Source: https://maestrix.ai/skills/create-competitive-intelligence > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a competitive intelligence system with monitoring, analysis, and response frameworks. ## CONTEXT NEEDED - Company name and description - Primary competitors (3-5) - Market/industry - Strategic priorities - Current competitive gaps (optional) ## COMPETITIVE INTELLIGENCE FRAMEWORK ### 1. Competitive Landscape Map **Direct Competitors** | Company | Overlap | Threat Level | Watch Priority | |---------|---------|--------------|----------------| **Indirect Competitors** | Company | Alternative Approach | Threat Level | Watch Priority | |---------|---------------------|--------------|----------------| **Emerging Threats** | Company | Why to Watch | Timeline | |---------|--------------|----------| ### 2. Market Position Matrix **Positioning Map** [Describe 2x2 matrix axes] - X-axis: [Dimension] - Y-axis: [Dimension] **Position by Competitor** | Competitor | X Position | Y Position | Strategy | |------------|------------|------------|----------| **White Space Opportunities** Underserved positions in the market ### 3. Competitive Dimension Analysis **Price Positioning** | Competitor | Price Level | Value Justification | |------------|-------------|---------------------| **Feature Depth** | Competitor | Breadth | Depth | Innovation Rate | |------------|---------|-------|-----------------| **Market Focus** | Competitor | Segments | Geography | Use Cases | |------------|----------|-----------|-----------| **Brand Perception** | Competitor | Positioning | Strengths | Weaknesses | |------------|-------------|-----------|------------| ### 4. Win/Loss Analysis Framework **Win Scenarios** | Competitor | When We Win | Why | Frequency | |------------|-------------|-----|-----------| **Loss Scenarios** | Competitor | When We Lose | Why | Frequency | |------------|--------------|-----|-----------| **Key Differentiators** What truly sets us apart ### 5. Intelligence Collection Plan **Primary Sources** | Source | What to Monitor | Frequency | Owner | |--------|-----------------|-----------|-------| **Secondary Sources** | Source | Value | Access Method | |--------|-------|---------------| **Human Intelligence** - Sales team debrief process - Customer feedback loops - Industry event insights - Partner intelligence ### 6. Monitoring Dashboard **Key Signals to Track** | Signal | Source | Threshold | Action | |--------|--------|-----------|--------| | Pricing change | Website | Any change | Review & respond | | New feature | Product | Major feature | Assess impact | | Funding | News | Any round | Strategy review | | Leadership | LinkedIn | C-level | Predict direction | | Hiring | Job boards | Volume spike | Identify focus | ### 7. Competitive Battlecards **Template per Competitor** **Quick Facts** - Founded / Funding / Employees - Key products - Target market - Recent news **Positioning** - How they position - vs. How we position **Strengths** [List with evidence] **Weaknesses** [List with evidence] **Objection Handling** | Their Claim | Truth | Our Response | |-------------|-------|--------------| **Trap Questions** Questions to ask that expose their weaknesses **When to Walk Away** Deals we shouldn't compete for ### 8. Response Playbooks **Price War Response** - When to match - When to hold - Value messaging **Feature Parity Response** - Acknowledge or ignore - Differentiation pivot - Timeline to match **Market Entry Response** - New competitor enters - Existing competitor expands - Disruptive new approach **Win-Back Campaigns** For lost customers ### 9. Competitive Content Strategy **Content Gaps vs. Competitors** | Topic | Competitor Coverage | Our Coverage | Action | |-------|---------------------|--------------|--------| **Comparison Content** - vs. pages needed - Feature comparison tables - Migration guides **Thought Leadership Opportunities** Where to own the conversation ### 10. Reporting & Communication **Weekly Intelligence Brief** - Key moves - Implications - Recommended actions **Monthly Competitive Review** - Trend analysis - Win/loss patterns - Strategic recommendations **Quarterly Strategy Session** - Landscape changes - Positioning adjustments - Resource allocation **Distribution** | Report | Audience | Format | Frequency | |--------|----------|--------|-----------| ## QUALITY RULES 1. NO opening comments 2. Actionable intelligence, not just data 3. Clear monitoring ownership 4. Response playbooks for key scenarios ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Battle card ↗' - sendPrompt: 'Create a competitive battle card for {{companyName}} — {{description}}, from this intelligence report' Secondary actions (neutral styling): - Label: 'Positioning ↗' sendPrompt: 'Update positioning for {{companyName}} — {{description}}, based on competitive insights' - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, with competitive differentiation' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, leveraging competitive advantages' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Content Angles Source: https://maestrix.ai/skills/create-content-angles > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Generate fresh, differentiated content angles that stand out from competitors. ## CONTEXT NEEDED - Company name and description - Content topic - Target audience - Existing content on topic (optional) - Competitor content (optional) ## CONTENT ANGLES FRAMEWORK ### 1. Topic Analysis **Current Coverage Assessment** - Common angles already covered - Overused approaches - Missing perspectives - Audience fatigue points ### 2. Angle Categories **Contrarian Angles** (5) Challenge conventional wisdom: - Angle + headline - Key argument - Evidence needed **Data-Driven Angles** (5) Lead with original research or analysis: - Angle + headline - Data source - Key insight **Story-Driven Angles** (5) Human narratives: - Angle + headline - Story type - Emotional hook **Expert Angles** (5) Authority perspectives: - Angle + headline - Expert type - Unique insight **Trend Angles** (5) Connect to current events: - Angle + headline - Trend connection - Timeliness factor ### 3. Differentiation Matrix | Angle | Novelty (1-10) | Credibility (1-10) | Engagement Potential | Difficulty | |-------|----------------|-------------------|---------------------|------------| ### 4. Deep Dive: Top 5 Angles For each selected angle: **Angle Title** **The Unique Perspective** What makes this different from existing content **Key Arguments** - Point 1 with evidence - Point 2 with evidence - Point 3 with evidence **Target Keyword** SEO optimization focus **Headline Options** (5) Different ways to frame it **Opening Hook** First paragraph that grabs attention **Content Structure** - Section breakdown - Key points per section - Data/proof needed **Visual Concepts** Supporting visuals to create **Distribution Strategy** Where this angle works best ### 5. Audience Segment Angles Same topic, different perspectives: | Audience | Their Concern | Angle Shift | Headline | |----------|---------------|-------------|----------| | Beginners | | | | | Advanced | | | | | Executives | | | | | Practitioners | | | | ### 6. Format-Angle Combinations | Angle | Best Format | Why | |-------|-------------|-----| | Contrarian | Op-ed, video essay | Strong POV | | Data | Report, infographic | Visual proof | | Story | Case study, interview | Human element | | Expert | Roundup, podcast | Multiple voices | | Trend | News analysis, social | Timeliness | ### 7. Cross-Platform Angle Adaptation **How to adapt the same angle for:** - Blog post - LinkedIn article - Twitter thread - YouTube video - Podcast episode - Email newsletter - Instagram carousel ### 8. Angle Testing Framework **Headline A/B Tests** - Variant A - Variant B - Test hypothesis **Content Performance Indicators** - Engagement signals - Share patterns - Comment quality - Conversion correlation ### 9. Angle Refresh Strategy For evergreen content: - When to revisit - How to update angle - New data to incorporate - Competitor movements to address ### 10. Content Brief Template For the winning angle, complete brief: - Working title - Target keyword - Unique angle statement - Key arguments - Proof points needed - Expert quotes needed - Visual requirements - Internal links - CTA ## QUALITY RULES 1. NO opening comments 2. Truly differentiated perspectives 3. Evidence requirements specified 4. Platform-appropriate adaptations ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Research facts ↗' - sendPrompt: 'Research facts for {{companyName}} — {{description}}, to validate and strengthen the content angles we just generated' Secondary actions (neutral styling): - Label: 'Blog article ↗' sendPrompt: 'Write a blog article for {{companyName}} — {{description}}, from the strongest angle' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, built around these content angles' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, using these content angles' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Content Distribution Source: https://maestrix.ai/skills/create-content-distribution > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Content Distribution Plan Most content fails not because it's bad — but because nobody sees it. This skill turns a single piece of content into a multi-channel distribution engine with platform-specific adaptations, timing, and amplification tactics. ## CONTEXT NEEDED - **Company name and description** - **Content to distribute**: Title, format, URL or summary of the content piece - **Content type**: Blog post / white paper / case study / video / report / infographic - **Target audience**: Who should see this content - **Available channels** *(optional)*: LinkedIn, Twitter/X, email, communities, paid, etc. - **Budget for paid amplification** *(optional)*: $0 / $100–500 / $500+ - **Goal**: Awareness / traffic / leads / engagement / backlinks --- ## Output Framework ### 1. Distribution Strategy Overview - Content asset summary (what we're distributing) - Primary goal and success metric - Distribution window (launch day + sustain period) - Channel priority ranking with rationale ### 2. Channel-by-Channel Plan For each channel (5–8 channels): **Channel Name** - **Format adaptation**: How to reshape the content for this platform - **Post copy**: Ready-to-publish text (platform-native tone and length) - **Visual requirements**: Image specs, carousel slides, video clips needed - **Hashtags / tags**: Platform-specific discovery tags - **Best posting time**: Day and time with reasoning - **Engagement tactic**: How to drive comments/shares (question, poll, hook) Channels to cover: - LinkedIn (post + article + comment strategy) - Twitter/X (thread + single tweet + quote tweet angles) - Email newsletter (dedicated send or inclusion) - Community/Reddit (where to post, how to frame) - Syndication (Medium, dev.to, industry publications) - YouTube/Video (if applicable — shorts, clips, full video) - Paid amplification (LinkedIn Ads, Meta boost, Google) - Internal/Employee advocacy (Slack share, team amplification) ### 3. Repurposing Matrix | Original Format | Derivative | Platform | Effort | |---|---|---|---| | Blog post | Twitter thread | Twitter/X | Low | | Blog post | LinkedIn carousel | LinkedIn | Medium | | Blog post | Email digest | Newsletter | Low | | Blog post | Infographic | All | High | | Blog post | Short video | TikTok/Reels | Medium | ### 4. Distribution Calendar | Day | Channel | Action | Content Variant | Status | |---|---|---|---|---| | Day 0 (Launch) | Blog | Publish original | Full article | — | | Day 0 | LinkedIn | Launch post | Hook + link | — | | Day 0 | Twitter/X | Thread | 5-tweet breakdown | — | | Day 1 | Email | Newsletter feature | Summary + CTA | — | | Day 2 | Reddit/Communities | Discussion post | Value-first framing | — | | Day 3 | LinkedIn | Follow-up post | Key stat or quote | — | | Day 7 | Twitter/X | Reminder tweet | Different angle | — | | Day 14 | Syndication | Republish | Medium/dev.to | — | | Day 30 | Email | Evergreen inclusion | Curated roundup | — | ### 5. Amplification Tactics - **Employee advocacy**: Template message for team to share - **Influencer seeding**: 3–5 people to tag or DM with the content - **Paid boost**: Budget allocation, targeting, expected reach - **Cross-promotion**: Partner newsletters, podcast mentions, guest posts - **SEO amplification**: Internal linking, backlink outreach targets ### 6. KPI Targets | Channel | Metric | Target | Measurement | |---|---|---|---| | Blog | Page views (30d) | — | Google Analytics | | LinkedIn | Impressions | — | LinkedIn Analytics | | Twitter/X | Thread views | — | Twitter Analytics | | Email | Open rate | — | ESP dashboard | | Email | Click rate | — | ESP dashboard | | Paid | Cost per click | — | Ad platform | | Overall | Leads generated | — | CRM / forms | ### 7. Post-Distribution Review Framework - What to measure at Day 7, Day 30, Day 90 - Content refresh triggers (when to update and re-distribute) - Learnings template for next distribution cycle --- ## Quality Rules - ✓ Every channel gets platform-native copy — no copy-paste across channels - ✓ Timing is specific (day + time), not vague - ✓ Repurposing creates genuinely new formats, not just re-links - ✓ Paid amplification is optional but planned - ✓ KPIs are measurable and channel-specific --- ## Example Prompts 1. "Create a distribution plan for our new blog post on 'AI in Marketing'" 2. "Distribute this case study across LinkedIn, email, and communities" 3. "Plan a 30-day distribution calendar for our white paper launch" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, optimized for the distribution plan we just created' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building on this distribution strategy' - Label: 'Email sequence ↗' sendPrompt: 'Create a lifecycle email sequence for {{companyName}} — {{description}}, to nurture leads from this content' - Label: 'Ad copy ↗' sendPrompt: 'Create ad copy for {{companyName}} — {{description}}, to amplify the content in this distribution plan' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Content Plan Source: https://maestrix.ai/skills/create-content-plan > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive content marketing distribution strategy with reader stories, brand voice, content themes, and a 20-day strategic calendar. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Keywords (optional) ## CONTENT PLAN DELIVERABLES ### 1. Reader Story "As a [target persona], I want [specific content solution] so that I can [achieve specific business outcome]." ### 2. Brand Voice Analysis Rate 1-100: Direct, Innovative, Supportive, Competent ### 3. Content Theme Matrix 5-7 core pillars with: - Theme name & focus - Target audience segment - Buyer journey stage - Content frequency - Key messages - Content examples ### 4. Content Curation Sources - Industry publications (5-7) - Thought leaders (10-15) - Competitors to monitor (3-5) - Trend monitoring tools ### 5. Content Type Distribution - Blog articles: X% - Video content: X% - Social media: X% - Infographics: X% - Case studies: X% - Webinars/Podcasts: X% - Email newsletters: X% ### 6. Engagement Metrics Framework - Awareness metrics (reach, traffic, followers) - Engagement metrics (rates, time on page) - Conversion metrics (leads, demos, SQLs) - Retention metrics (repeat engagement) ### 7. 20-Day Strategic Content Calendar Daily entries with: - Content idea - Platform - Source type - Target metrics - Cross-promotion plan - CTA strategy - Hashtags/keywords ## QUALITY RULES 1. NO opening comments 2. Actionable, specific recommendations 3. Measurable goals for each metric ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Generate first article ↗' - sendPrompt: 'Write a blog article for {{companyName}} — {{description}}, based on the top priority topic from this content plan' Secondary actions (neutral styling): - Label: 'Social posts ↗' sendPrompt: 'Create social media posts for {{companyName}} — {{description}}, based on the content pillars from this plan' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, aligned with the content themes we just defined' - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, building on this content plan' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Ecosystem Messaging Source: https://maestrix.ai/skills/create-ecosystem-messaging > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create messaging for partner ecosystem, integrations, and alliance marketing. ## CONTEXT NEEDED - Company name and description - Integration/partner type - Partner name(s) if specific - Target audience for integration - Use cases enabled (optional) ## ECOSYSTEM MESSAGING FRAMEWORK ### 1. Integration Value Proposition **Combined Value Statement** [Your product] + [Partner product] = [Unique outcome neither achieves alone] **Integration Headline** - Benefit-focused (3 options) - Problem-focused (3 options) - Workflow-focused (3 options) ### 2. Integration Story Arc **Before Integration** - Current workflow pain points - Manual steps required - Data silos - Time/cost impact **The Connection** - What data/actions flow between systems - Technical simplicity **After Integration** - New streamlined workflow - Time saved - Errors eliminated - New capabilities unlocked ### 3. Use Case Messaging For each use case (4-6): - Use case name - Target persona - Trigger scenario - How integration helps - Quantified outcome - Setup complexity ### 4. Partner Type Messaging Templates **Technology Partner** - Joint value prop - Technical benefits - Implementation ease - Combined differentiators **Solution Partner** - Service value-add - Expertise areas - Customer success focus **Marketplace Partner** - Discovery benefits - Trust signals - Comparison points ### 5. Audience-Specific Messaging **For Partner's Customers** - Why add your product - Easy onboarding message - Risk reversal **For Your Customers** - Why connect partner - Enhanced capability message - Quick setup promise **For Joint Prospects** - Combined solution story - Best-of-both message - Single vendor simplicity ### 6. Social Proof for Integrations - Joint customer quotes - Usage statistics - Implementation success rates - Time-to-value metrics ### 7. Competitive Positioning **vs. Native Features** - Specialist advantage - Flexibility benefit **vs. Alternative Integrations** - Depth of integration - Ease of setup - Ongoing reliability ### 8. Channel Assets **Partner Directory Listing** - Title (60 chars) - Short description (150 chars) - Full description (300 words) - Key features (5 bullets) - Categories/tags **Co-Marketing Email** - Subject lines (3) - Preview text - Body copy - CTAs **Social Announcement** - Twitter/X - LinkedIn - Visual concepts ### 9. Sales Enablement **Partner Pitch Deck Slides** - Integration overview - Use cases - Customer proof - Next steps **Battle Card Content** - Quick integration facts - Key selling points - Common questions - Competitive comparison ### 10. Launch Sequence **Pre-Launch**: Tease to existing users **Launch Day**: Joint announcement **Week 1**: Deep-dive content **Week 2**: Customer spotlight **Ongoing**: Use case series ## QUALITY RULES 1. NO opening comments 2. Mutual value emphasis (not one-sided) 3. Specific use cases with outcomes 4. Easy-to-adapt templates ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a partner/integrations page for {{companyName}} — {{description}}, using this ecosystem messaging' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create partner outreach emails for {{companyName}} — {{description}}, based on this ecosystem strategy' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting the ecosystem' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, with partner co-marketing topics' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Event Marketing Source: https://maestrix.ai/skills/create-event-marketing > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create comprehensive event marketing strategy with promotion, execution, and follow-up plans. ## CONTEXT NEEDED - Company name and description - Event type (webinar, conference, workshop, etc.) - Event topic and value proposition - Target audience - Event date (optional) ## EVENT MARKETING FRAMEWORK ### 1. Event Strategy Overview **Event Type**: [Webinar/Conference/Workshop/etc.] **Theme**: [Core topic/title] **Value Proposition**: [Why attend in one sentence] **Target Audience**: [Who should attend] **Registration Goal**: [Number] **Attendance Goal**: [Number, typically 40-50% of registrations] **Conversion Goal**: [Post-event action] ### 2. Event Positioning **Title Options** (5) - Benefit-focused - Outcome-focused - Problem-focused - Curiosity-focused - Expert-focused **Description** (150 words) What attendees will learn and gain **Key Takeaways** (5) Bullet points for promotion ### 3. Promotional Timeline **6 Weeks Out** - Landing page live - Initial email announcement - Social tease **4 Weeks Out** - Speaker announcements - Partner promotions - Paid campaign launch **2 Weeks Out** - Reminder campaign - Content previews - Urgency messaging **1 Week Out** - Final push - "Last chance" messaging - Calendar reminders **Day Of** - Morning reminder - 1-hour reminder - Live engagement **Post-Event** - Thank you email - Recording access - Nurture sequence ### 4. Landing Page Copy **Headline**: [Event title] **Subhead**: [Key benefit] **Body Copy**: [Value proposition expansion] **What You'll Learn Section** - Takeaway 1 - Takeaway 2 - Takeaway 3 - Takeaway 4 - Takeaway 5 **Speaker Bios** [Name, title, expertise, photo] **Social Proof** Previous attendee testimonials or metrics **CTA Button Text**: [Register Now / Save Your Spot] ### 5. Email Sequence **Email 1: Announcement** - Subject lines (5) - Preview text - Body copy - CTA **Email 2: Value Reminder** - Subject lines (3) - Key takeaways focus - CTA **Email 3: Speaker Spotlight** - Subject lines (3) - Speaker credibility - CTA **Email 4: Social Proof** - Subject lines (3) - Registration numbers - Testimonials - CTA **Email 5: Urgency** - Subject lines (3) - Limited spots/time - CTA **Email 6: Day-Of Reminder** - Subject line - Join link - Quick value reminder ### 6. Social Media Content **LinkedIn Posts** (5) **Twitter Posts** (10) **Instagram Posts** (3) **Facebook Posts** (3) Include: - Announcement post - Speaker highlight - Countdown posts - Live engagement prompts - Recording promotion **Event Hashtag**: #[Suggestion] ### 7. Partner/Speaker Promotion Kit **Swipe Copy** - Email template - Social posts - SMS/Slack messages **Visual Assets Needed** - Event banner - Speaker cards - Social graphics ### 8. Event Content **Presentation Outline** - Opening hook (3 min) - Agenda overview - Main content blocks - Interactive elements (polls, Q&A) - Key takeaways recap - CTA/next steps - Q&A **Engagement Tactics** - Polls - Chat prompts - Q&A strategy - Live reactions ### 9. Follow-Up Sequence **Attended Sequence** - Email 1: Thank you + recording + resources - Email 2: Key insights recap + offer - Email 3: Related content + CTA - Email 4: Personal outreach/demo offer **No-Show Sequence** - Email 1: Recording access - Email 2: Key highlights - Email 3: Next event/offer ### 10. Metrics & Success Criteria | Metric | Target | Tracking | |--------|--------|----------| | Registrations | | | | Attendance rate | 40-50% | | | Engagement (chat/polls) | | | | Recording views | | | | Post-event conversions | | | | Pipeline generated | | | ## QUALITY RULES 1. NO opening comments 2. Complete promotional timeline 3. Multi-channel content 4. Clear conversion path ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this event' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create event invitation emails for {{companyName}} — {{description}}' - Label: 'Landing page ↗' sendPrompt: 'Create an event landing page for {{companyName}} — {{description}}' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, around this event' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Faqs Source: https://maestrix.ai/skills/create-faqs > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create comprehensive FAQ content that answers questions, handles objections, and improves SEO. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Common customer questions (optional) - Competitor FAQs (optional) ## FAQ FRAMEWORK ### 1. Question Categories **Product/Service Questions** (10-15) About features, capabilities, and how it works **Pricing & Plans Questions** (5-10) About cost, billing, and plan differences **Getting Started Questions** (5-10) About onboarding, setup, and first steps **Technical Questions** (5-10) About integrations, requirements, and specs **Comparison Questions** (5-7) About alternatives and competition **Trust & Security Questions** (5-7) About data, compliance, and reliability **Support Questions** (5-7) About help, training, and resources ### 2. Complete FAQ Set For each question: **Question**: [Natural language question as users would ask] **Short Answer**: [1-2 sentence direct response] **Expanded Answer**: [2-3 paragraph comprehensive response] - Context and detail - How-to elements if applicable - Links to related resources **Related Questions**: [2-3 connected FAQs] ### 3. Objection-Handling FAQs Convert sales objections to FAQs: | Objection | FAQ Question | Answer Strategy | |-----------|--------------|-----------------| | Too expensive | "How does pricing compare to [alternative]?" | Value justification | | Too complex | "How long does setup take?" | Simplicity proof | | Not ready | "Can I try before committing?" | Risk reversal | | Need X feature | "Do you integrate with X?" | Capability proof | | Not sure it works | "What results do customers see?" | Social proof | ### 4. SEO-Optimized FAQs **Target Keywords** | Keyword | Question | Search Volume | |---------|----------|---------------| **People Also Ask** (10 questions) Questions from Google's PAA boxes **Long-Tail Queries** (10 questions) Specific, lower-competition questions ### 5. Schema Markup Ready Format for FAQ schema: ```json { "@type": "FAQPage", "mainEntity": [ { "@type": "Question", "name": "[Question]", "acceptedAnswer": { "@type": "Answer", "text": "[Answer]" } } ] } ``` ### 6. Tone & Voice Guidelines **Answer Style** - Direct (answer first, context second) - Helpful (not defensive) - Clear (no jargon) - Action-oriented (include next steps) **Do's** - Acknowledge the concern - Provide specific answers - Link to resources - Offer alternatives **Don'ts** - Avoid defensive language - Don't redirect without answering - Don't use corporate speak - Don't leave questions partially answered ### 7. FAQ Page Structure **Organization Options** - By category (recommended for 20+) - By user type - By journey stage - By popularity **UX Elements** - Search functionality - Expandable accordions - Jump links - Related content - Contact option ### 8. Maintenance Plan **Review Cadence** - Weekly: Check for new common questions - Monthly: Update answers with new info - Quarterly: Audit for accuracy - Annually: Full content review **Sources for New FAQs** - Support tickets - Chat logs - Sales call notes - Search analytics - Social mentions ### 9. Channel Adaptations **Chatbot Responses** Shorter, conversational versions **Email Templates** For support responses **Social Media Responses** Brief, linkable answers ### 10. Measurement **FAQ Performance Metrics** | Metric | Target | Tracking Method | |--------|--------|-----------------| | Page views | | Analytics | | Search usage | | Site search | | Time on page | | Analytics | | Support ticket reduction | | Support system | | Exit rate | | Analytics | ## QUALITY RULES 1. NO opening comments 2. Answer the question directly first 3. Natural language (how users actually ask) 4. Include next steps and links ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, incorporating these FAQs' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, turning FAQ topics into articles' - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, targeting FAQ-related search queries' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, addressing these common questions' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Feature Messaging Source: https://maestrix.ai/skills/create-feature-messaging > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create compelling feature-specific messaging that connects capabilities to customer outcomes. ## CONTEXT NEEDED - Company name and description - Feature name and description - Target audience for this feature - Competitive alternatives (optional) - Customer use cases (optional) ## FEATURE MESSAGING FRAMEWORK ### 1. Feature Positioning Statement For [target user] who [specific need], [feature name] is a [category] that [primary benefit]. Unlike [alternative approaches], it [key differentiator]. ### 2. Feature-Benefit Ladder | Feature Aspect | Capability | Benefit | Outcome | |----------------|------------|---------|---------| | [What it is] | [What it does] | [What user gets] | [Business result] | Create 5-7 complete ladders. ### 3. Headline Options **Benefit-Led Headlines** (5) - Focus on outcome **Problem-Led Headlines** (5) - Focus on pain solved **Differentiator Headlines** (5) - Focus on unique approach ### 4. Short-Form Descriptions **One-liner** (10 words max) **Elevator pitch** (30 words) **Feature card** (50 words) **Full description** (100 words) ### 5. Use Case Messaging For each use case (3-5): - Use case name - Target persona - Situation/trigger - How feature helps - Outcome achieved - Proof point ### 6. Objection Responses | Objection | Response | Proof Point | |-----------|----------|-------------| Address 5-7 common objections. ### 7. Competitive Positioning **vs. Manual Approach** - Their way: [pain] - Our way: [benefit] **vs. Competitor A** - Their approach: [limitation] - Our approach: [advantage] **vs. Competitor B** - Their approach: [limitation] - Our approach: [advantage] ### 8. Social Proof Integration - Customer quote templates - Metric callouts - Case study hooks - Usage statistics ### 9. Channel-Specific Messaging **Product UI** - Tooltip text - Empty state messaging - Onboarding copy - Upgrade prompts **Marketing Website** - Hero section - Feature section - Comparison table **Email** - Announcement subject lines - Feature highlight snippet **Social Media** - Twitter/X post - LinkedIn post - Visual hook concept ### 10. Launch Messaging Sequence **Day 1**: Announcement **Week 1**: Deep dive **Week 2**: Use cases **Week 3**: Customer stories **Week 4**: Comparison content ## QUALITY RULES 1. NO opening comments 2. Benefits before features always 3. Specific outcomes with metrics 4. Persona-appropriate language ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a feature landing page for {{companyName}} — {{description}}, using this feature messaging' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, promoting these features' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, highlighting key features' - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, demoing these features' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Google Ads Source: https://maestrix.ai/skills/create-google-ads > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive Google Ads campaign plan focusing on driving sales and conversions with three distinct campaigns targeting TOFU (awareness), MOFU (consideration), and BOFU (conversion) stages. ## CONTEXT NEEDED To generate Google Ads, the user should provide: - Company name and description - Products or services offered - Target audience - Keywords or themes (optional) - Target persona (optional - for personalized messaging) ## CAMPAIGN STRUCTURE Develop **3 campaigns** (TOFU, MOFU, BOFU) with **3 ad groups each** (9 total ad groups). ### For Each Ad Group, Provide: 1. **Ad Group Name**: Clear name with purpose and campaign stage (e.g., "TOFU - Problem Awareness - [Theme]") 2. **Keywords** (exactly 10 per ad group): - NO overlapping keywords across ad groups or campaigns - Use persona's language and search terms if provided - Mix of match types appropriate for the stage 3. **Headlines** (exactly 15 per ad group): - Minimum 25 characters, strict maximum 30 characters - NO commas (Google Ads restriction) - Use keywords and rhetorical devices - Align with buyer's journey intent for the campaign stage - Apply copywriting hooks and formulas: - Problem-solution - Before-after - Social proof - Urgency - Curiosity gaps 4. **Descriptions** (exactly 4 per ad group): - Minimum 75 characters, strict maximum 90 characters - NO commas - Use keywords and rhetorical devices - Align with buyer's journey intent - Address persona's motivations and overcome barriers 5. **Sitelinks** (2-3 per ad group): - Text (max 25 characters) - Description (max 35 characters) - Address common objections 6. **Callouts** (3-5 per ad group): - Max 25 characters each - Highlight benefits relevant to the persona ## OUTPUT FORMAT Output as valid JSON with this structure: ```json { "campaigns": [ { "name": "TOFU - Awareness Campaign", "objective": "Brand awareness and problem recognition", "adGroups": [ { "name": "TOFU - [Theme] - Ad Group 1", "keywords": ["keyword 1", "keyword 2", "...10 total"], "titles": ["Headline 1", "Headline 2", "...15 total"], "descriptions": ["Description 1 (75-90 chars)", "...4 total"], "sitelinks": [ {"text": "Sitelink Text", "description": "Short description"}, {"text": "Sitelink 2", "description": "Description 2"} ], "callouts": ["Callout 1", "Callout 2", "Callout 3"] } ] }, { "name": "MOFU - Consideration Campaign", "objective": "Solution evaluation and comparison", "adGroups": [...] }, { "name": "BOFU - Conversion Campaign", "objective": "Purchase decision and action", "adGroups": [...] } ] } ``` ## CAMPAIGN STAGE GUIDANCE ### TOFU (Top of Funnel) - Awareness - **Keywords**: Informational, problem-focused - **Headlines**: Problem awareness, industry challenges - **Tone**: Educational, thought-provoking - **CTAs**: Learn, Discover, See How ### MOFU (Middle of Funnel) - Consideration - **Keywords**: Solution-seeking, comparison - **Headlines**: Solution benefits, differentiation - **Tone**: Helpful, comparative - **CTAs**: Compare, Explore, Get Demo ### BOFU (Bottom of Funnel) - Conversion - **Keywords**: Transactional, brand + action - **Headlines**: Offers, guarantees, social proof - **Tone**: Urgent, confident - **CTAs**: Start Free, Buy Now, Get Started ## QUALITY RULES 1. **NO opening comments** - Start directly with the JSON 2. **NO text before or after the JSON** - Pure JSON output 3. **Valid JSON format** - Properly structured and parseable 4. **Exact counts**: 10 keywords, 15 headlines, 4 descriptions per ad group 5. **Character limits are strict** - Google will reject non-compliant ads 6. **No commas in headlines/descriptions** - Google Ads restriction 7. **No duplicate keywords** across any ad groups 8. **Headlines must be unique** - Don't repeat across ad groups 9. **Persona-specific language** if persona context provided ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Meta Ads ↗' - sendPrompt: 'Create Meta Ads for {{companyName}} — {{description}}, complementing the Google Ads we just created' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a destination for these Google Ads' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to optimize these ad campaigns' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, reinforcing the ad messaging' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Gtm Motion Source: https://maestrix.ai/skills/create-gtm-motion > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Analyze your business to determine the optimal GTM motion and create an implementation plan. ## CONTEXT NEEDED - Company name and description - Product type and pricing - Average contract value (ACV) - Sales cycle length - Current team size - Target customer profile ## GTM MOTION ASSESSMENT ### 1. Motion Fit Analysis Score each motion (1-5) based on your business characteristics: **Product-Led Growth (PLG)** | Factor | Score (1-5) | Notes | |--------|-------------|-------| | Self-serve possible | | Can users get value without sales? | | Low ACV (<$5K) | | Price point supports self-serve | | Viral/network potential | | Users invite others naturally | | Simple onboarding | | Time-to-value under 10 minutes | | Freemium viable | | Free tier economics work | | **PLG Total** | /25 | | **Sales-Led Growth (SLG)** | Factor | Score (1-5) | Notes | |--------|-------------|-------| | High ACV (>$25K) | | Justifies sales cost | | Complex buyer journey | | Multiple stakeholders | | Customization needed | | Solution requires configuration | | Long sales cycle (>30 days) | | Relationship building matters | | Enterprise target | | Security, compliance, legal reviews | | **SLG Total** | /25 | | **Marketing-Led Growth (MLG)** | Factor | Score (1-5) | Notes | |--------|-------------|-------| | Strong content opportunity | | Thought leadership viable | | High search volume | | Prospects actively searching | | Long consideration phase | | Nurturing adds value | | Brand differentiation | | Story matters to buyers | | Community potential | | Audience wants to congregate | | **MLG Total** | /25 | | **Partner-Led Growth (PartnerLG)** | Factor | Score (1-5) | Notes | |--------|-------------|-------| | Ecosystem dependency | | Works better with other tools | | Channel efficiency | | Partners have customer access | | Implementation services | | Requires specialist deployment | | Industry specialization | | Vertical expertise needed | | Geographic expansion | | Partners provide local presence | | **PartnerLG Total** | /25 | | **Founder-Led Growth (FLG)** | Factor | Score (1-5) | Notes | |--------|-------------|-------| | Early stage (<$1M ARR) | | Still finding PMF | | Founder domain expertise | | Deep industry credibility | | High-touch sales required | | Conviction selling needed | | Network-driven market | | Relationships unlock deals | | Learning mode | | Need direct customer feedback | | **FLG Total** | /25 | | ### 2. Recommended Motion **Primary Motion:** [Highest scoring] **Secondary Motion:** [Second highest, if within 5 points] **Motion Mix Ratio:** [e.g., 70% PLG / 30% SLG] ### 3. Motion-Specific Playbook #### If Product-Led (PLG) **Funnel Structure** ``` Website Visitor → Sign Up → Activation → Paid → Expansion ``` **Key Investments** | Area | Priority | Resources Needed | |------|----------|------------------| | Self-serve onboarding | Critical | Product, UX | | In-app analytics | Critical | Data, Product | | Free tier optimization | High | Product, Growth | | Virality features | Medium | Product | | PLG content (docs, tutorials) | High | Marketing | **Metrics Focus** - Activation rate - Time to value - PQL (Product Qualified Lead) conversion - Expansion revenue % #### If Sales-Led (SLG) **Funnel Structure** ``` Lead → MQL → SQL → Opportunity → Closed-Won ``` **Key Investments** | Area | Priority | Resources Needed | |------|----------|------------------| | SDR team | Critical | Sales | | CRM/Sales tools | Critical | Ops | | Sales enablement | High | Marketing, Sales | | Demo environment | High | Product | | Account executives | Critical | Sales | **Metrics Focus** - SQL-to-close rate - Average deal size - Sales cycle length - Win rate vs competitors #### If Marketing-Led (MLG) **Funnel Structure** ``` Content → Lead → Nurtured → Sales Ready → Customer ``` **Key Investments** | Area | Priority | Resources Needed | |------|----------|------------------| | Content engine | Critical | Marketing | | SEO/Organic | Critical | Marketing, Content | | Marketing automation | High | Marketing Ops | | Community building | Medium | Marketing | | Events/Webinars | Medium | Marketing | **Metrics Focus** - Organic traffic growth - Content-attributed pipeline - MQL-to-SQL conversion - CAC by channel #### If Partner-Led (PartnerLG) **Funnel Structure** ``` Partner Referral → Joint Qualification → Co-Sell → Implementation ``` **Key Investments** | Area | Priority | Resources Needed | |------|----------|------------------| | Partner program | Critical | Partnerships | | Integration development | Critical | Engineering | | Partner enablement | High | Marketing, Sales | | Co-marketing | Medium | Marketing | | Partner portal | Medium | Product | **Metrics Focus** - Partner-sourced pipeline - Partner-influenced revenue - Partner activation rate - Co-sell win rate ### 4. Resource Requirements | Role | Motion Fit | When to Hire | Cost Range | |------|------------|--------------|------------| ### 5. Tech Stack by Motion **PLG Stack** - Product analytics (Amplitude, Mixpanel) - Onboarding (Appcues, Pendo) - Billing (Stripe, Chargebee) - In-app messaging (Intercom) **SLG Stack** - CRM (Salesforce, HubSpot) - Sales engagement (Outreach, Salesloft) - Revenue intelligence (Gong, Chorus) - CPQ (DealHub, PandaDoc) **MLG Stack** - Marketing automation (HubSpot, Marketo) - CMS (Webflow, WordPress) - SEO tools (Ahrefs, Semrush) - Community (Discourse, Circle) ### 6. Transition Planning If your current motion differs from recommended: **Phase 1: Foundation (0-3 months)** - Key hires - Tool implementation - Process documentation **Phase 2: Pilot (3-6 months)** - Small-scale testing - Metric benchmarking - Iteration cycles **Phase 3: Scale (6-12 months)** - Full rollout - Team expansion - Motion optimization ### 7. Hybrid Motion Design If running multiple motions: | Segment | Motion | Trigger | Handoff Point | |---------|--------|---------|---------------| | SMB | PLG | Self-serve signup | Usage threshold | | Mid-Market | PLG + SLG | Expansion signal | Sales outreach | | Enterprise | SLG | Inbound demo request | AE assignment | ## QUALITY RULES 1. NO opening comments 2. Score all five motions before recommending 3. Resource estimates must be realistic 4. Include transition plan if changing motions 5. Specify hybrid triggers if multiple motions ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Channel matrix ↗' - sendPrompt: 'Create a channel matrix for {{companyName}} — {{description}}, aligned with this GTM motion' Secondary actions (neutral styling): - Label: 'Pipeline playbook ↗' sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, for this GTM motion' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, supporting this GTM motion' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, aligned with this motion' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Gtm Strategy Source: https://maestrix.ai/skills/create-gtm-strategy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive go-to-market strategy for product/feature launch or market entry. ## CONTEXT NEEDED - Company name and description - Product/feature being launched - Target audience - Launch timeline - Budget range (optional) - Competitive landscape (optional) ## GTM STRATEGY FRAMEWORK ### 1. Market Opportunity Summary - Total Addressable Market (TAM) - Serviceable Addressable Market (SAM) - Serviceable Obtainable Market (SOM) - Market timing factors - Competitive white space ### 2. Target Audience Definition **Primary Segment** - Profile description - Size and accessibility - Pain points addressed - Willingness to pay - Decision-making process **Secondary Segment** - Profile description - Expansion potential **Anti-Personas** (who NOT to target initially) ### 3. Value Proposition & Positioning **Positioning Statement** For [target customer] who [need], [product] is a [category] that [key benefit]. Unlike [alternatives], we [key differentiator]. **Value Pillars** (3-4): - Pillar name - Customer benefit - Proof point - Competitive advantage ### 4. Pricing Strategy | Tier | Price Point | Target Segment | Key Features | Rationale | |------|-------------|----------------|--------------|-----------| - Pricing model recommendation - Competitive price positioning - Promotional pricing (if applicable) ### 5. Channel Strategy **Primary Channels** | Channel | Role | Investment Level | Expected CAC | Timeline to Results | |---------|------|------------------|--------------|---------------------| **Channel Mix Rationale** ### 6. Launch Phases **Phase 1: Pre-Launch** (Weeks -4 to 0) - Activities - Goals - Deliverables **Phase 2: Launch** (Weeks 0-2) - Activities - Goals - Deliverables **Phase 3: Post-Launch** (Weeks 3-12) - Activities - Goals - Deliverables ### 7. Content & Messaging Plan | Asset | Purpose | Audience | Channel | Due Date | |-------|---------|----------|---------|----------| List 15-20 launch assets. ### 8. Sales Enablement - Battle cards - Demo scripts - Objection handling - Competitive positioning - ROI calculator ### 9. Partnership & Ecosystem - Strategic partners to engage - Integration opportunities - Co-marketing potential ### 10. Metrics & Success Criteria **Launch Metrics** (Day 1-30) - Metric, target, tracking method **Growth Metrics** (Day 31-90) - Metric, target, tracking method **Efficiency Metrics** (Ongoing) - CAC, LTV, payback period targets ### 11. Risk Mitigation | Risk | Probability | Impact | Mitigation Strategy | |------|-------------|--------|---------------------| ### 12. Budget Allocation | Category | % of Budget | Activities | |----------|-------------|------------| ## QUALITY RULES 1. NO opening comments 2. Specific, actionable tactics 3. Realistic timelines and metrics 4. Clear ownership and accountability ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, aligned with the GTM strategy we just built' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, targeting the ICP segments from this GTM strategy' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, based on the channel strategy from this GTM plan' - Label: 'Pipeline playbook ↗' sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, building on this GTM strategy' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Icp Sizing Source: https://maestrix.ai/skills/create-icp-sizing > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a detailed Ideal Customer Profile (ICP) with market sizing and qualification criteria. ## CONTEXT NEEDED - Company name and description - Products or services offered - Current customer examples (optional) - Industry focus (optional) - Geographic scope ## ICP FRAMEWORK ### 1. ICP Definition Summary **One-Line ICP Statement** [Company type] with [characteristics] that [situation/trigger] and value [outcome]. ### 2. Firmographic Criteria | Attribute | Ideal | Acceptable | Disqualifying | |-----------|-------|------------|---------------| | Industry/Vertical | | | | | Company Size (employees) | | | | | Revenue Range | | | | | Geography | | | | | Growth Stage | | | | | Tech Stack | | | | | Business Model | | | | ### 3. Technographic Profile - Required technologies - Complementary tools - Technology maturity level - IT infrastructure requirements - Integration dependencies ### 4. Behavioral Indicators **Buying Signals** - Active triggers (what they're doing now) - Passive indicators (characteristics that correlate) - Negative signals (disqualifiers) **Engagement Patterns** - Content consumption habits - Community participation - Event attendance - Vendor evaluation process ### 5. Organizational Characteristics - Decision-making structure - Budget ownership - Procurement process - Typical buying committee - Champion profile - Economic buyer profile ### 6. Problem-Solution Fit | Pain Point | Severity (1-10) | Current Solution | Switching Cost | Our Advantage | |------------|-----------------|------------------|----------------|---------------| ### 7. Market Sizing **Total Addressable Market (TAM)** - Calculation methodology - Data sources - Total market value **Serviceable Addressable Market (SAM)** - Geographic/segment filters applied - Resulting market size **Serviceable Obtainable Market (SOM)** - Realistic capture rate - Competitive factors - 3-year target ### 8. ICP Tiers **Tier 1: Perfect Fit** (Score 90-100) - Characteristics - % of market - Estimated count - Recommended approach **Tier 2: Good Fit** (Score 70-89) - Characteristics - % of market - Estimated count - Recommended approach **Tier 3: Qualified** (Score 50-69) - Characteristics - % of market - Estimated count - Recommended approach ### 9. Scoring Model | Criterion | Weight | Score 1 | Score 3 | Score 5 | |-----------|--------|---------|---------|---------| Total possible score: 100 ### 10. ICP Validation Checklist 10-point checklist for validating prospects against ICP. ### 11. Expansion Signals Indicators that a current customer is ready for: - Upsell - Cross-sell - Expansion to other teams/departments ## QUALITY RULES 1. NO opening comments 2. Quantified criteria (not vague) 3. Realistic market sizing with sources 4. Actionable scoring model ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** A **Layout spec:** Hero card (avatar initials, name, title, summary) → stat chip row (4 metrics) → 2-column card grid (Goals vs Pain Points, Motivations vs KPIs) → tag cloud (sources/tools) → quote row (3 VoC cards) → 2-column highlight (risk --warning, opportunity --success) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Persona ↗' - sendPrompt: 'Generate a detailed persona for {{companyName}} — {{description}}, based on the ICP sizing we just completed' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, targeting the primary ICP segment' - Label: 'Positioning ↗' sendPrompt: 'Create positioning for {{companyName}} — {{description}}, tailored to the ICP we just sized' - Label: 'GTM strategy ↗' sendPrompt: 'Create a GTM strategy for {{companyName}} — {{description}}, focused on the market segments identified' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Influencer Campaign Source: https://maestrix.ai/skills/create-influencer-campaign > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system. Never use CSS variables. Never make a secondary API call. Design a comprehensive influencer marketing strategy with creator tier mix, platform priorities, outreach scripts, content briefs, vetting criteria, budget allocation, and ROI measurement. ## CONTEXT NEEDED - Company name and description - Campaign goal (awareness, launch, UGC, conversions, community, event) - Target influencer tier (nano / micro / mid-tier / macro / mega) - Target platforms (Instagram, TikTok, YouTube, LinkedIn, Twitter) - Budget (optional) - Target audience ## INFLUENCER CAMPAIGN FRAMEWORK ### 1. Executive Summary Campaign goal, primary platforms, tier focus, recommended # of influencers, expected reach, estimated EMV, and a 3-4 sentence strategic rationale. ### 2. Influencer Tier Strategy Tier comparison table (nano / micro / mid-tier / macro / mega) with followers, cost range, engagement rate, best use, and priority for this campaign. Recommended budget mix across tiers. ### 3. Platform Strategy Per-platform priority (🔴/🟡/🟢), content type, audience match, best format. Platform-specific tactics for Instagram, TikTok, YouTube, Twitter/X, LinkedIn. ### 4. Content Brief Templates - Sponsored post brief (campaign name, key messages, must-haves, must-avoids, tone) - Story/Reel brief (hook → middle → CTA) ### 5. Influencer Vetting Criteria Selection scorecard (audience match, engagement rate, content quality, brand alignment, authenticity, professionalism). Red flags to avoid. 4-step vetting process. ### 6. Budget Allocation Budget framework (or budget-size guide if no budget given). Cost benchmarks by platform × tier (Instagram, TikTok, YouTube). ### 7. Campaign Structure Full timeline (planning → outreach → production → launch → optimization → analysis) and 3 campaign types (always-on, product launch blitz, UGC campaign). ### 8. Measurement Framework KPIs by goal (awareness, engagement, traffic, conversions). Attribution methods (unique codes, UTMs, affiliate links, brand lift). Weekly + campaign reporting templates. ### 9. Outreach Templates - Initial outreach DM - Follow-up email with compensation, deliverables, timeline ### 10. Legal & Compliance FTC disclosure requirements, platform-specific labels, contract essentials checklist. ### 11. Implementation Checklist Pre / during / post campaign action items. ## QUALITY RULES 1. NO opening fluff 2. Specific platform and tier recommendations 3. Ready-to-use outreach templates 4. Realistic budget benchmarks 5. Clear measurement framework ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract **Pattern:** B (Tabbed Dashboard) Persistent header (campaign goal + tier + platforms + EMV estimate as stat chips) → tab bar (Strategy → Tiers → Platforms → Outreach → Briefs → Measurement) → per-tab content → attribution footer → What's Next action bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ### What's next actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this influencer campaign to my {{companyName}} project" - **create_social_posts** (green, primary) — sendPrompt: "Create social posts for {{companyName}} to amplify the influencer campaign we just designed" - **create_content_distribution** (neutral, secondary) — sendPrompt: "Create a content distribution plan for {{companyName}} around the influencer campaign" - **create_ad_copy** (neutral, secondary) — sendPrompt: "Create whitelist / dark post ad copy for {{companyName}} based on the influencer briefs" - **generate_persona** (neutral, secondary) — sendPrompt: "Generate a target persona for {{companyName}} to refine influencer audience match" --- ## Skill: Create Landing Page Source: https://maestrix.ai/skills/create-landing-page > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create the most comprehensive, professional copywriting strategy document for a high-converting landing page using psychological frameworks that drive massive action and revenue. ## CONTEXT NEEDED To generate landing page copy, the user should provide: - Company name and description - Products or services offered - Target audience or persona - Value propositions (optional) - Competitors (optional) ## PSYCHOLOGICAL FRAMEWORKS TO APPLY - **StoryBrand 7-part framework** (Customer Hero's Journey) - **PAS** (Problem-Agitation-Solution) with emotional amplification - **Before-After-Bridge** with transformation storytelling - **Social proof hierarchy** and credibility stacking - **Scarcity and urgency** psychology with ethical persuasion - **Loss aversion** and risk reversal mastery - **Authority positioning** and expertise demonstration - **Cognitive biases**: anchoring, reciprocity, social proof, commitment/consistency ## LANDING PAGE SECTIONS ### 🎯 Hero Section **Headline Strategy** - Pre-headline hook (curiosity/urgency) - Main headline (primary benefit with emotional impact) - Subheadline (supporting details with specificity) - Visual concept (hero image/video description) - Social proof snippet (trust indicator or impressive stat) - Primary CTA (action-oriented button text) - Risk reversal (guarantee or risk-free offer) ### ⚡ Problem & Agitation **Pain Point Amplification** - Hook story (1-sentence relatable scenario) - Core pain points (3-5 specific pains) - Consequences (what happens if unsolved) - Emotional trigger (deep emotional consequence) ### 💡 Solution Introduction **Unique Mechanism** - Solution title (compelling solution name) - Unique mechanism (how your solution is different) - Big promise (primary transformation promise) - Proof elements (evidence that supports claims) ### 🔄 How It Works **Simple 3-Step Process** 1. Step 1: [Title] - [Description] → [Benefit] 2. Step 2: [Title] - [Description] → [Benefit] 3. Step 3: [Title] - [Description] → [Benefit] ### ⭐ Social Proof Stack **Credibility Building** - Customer Success Stories (3-5 testimonials with results) - Trust Statistics (impressive numbers) - Trust Logos (companies/publications/certifications) ### 🚀 Benefits Breakdown **Transformation Outcomes** For each major benefit (3-5): - Benefit title - What it means - Specific outcome - Proof point ### 🔧 Features Showcase **Powerful Capabilities** For each feature (5-8): - Feature name - Technical capability - User benefit translation ### 🆚 Why Choose Us **Competitive Advantages** Comparison table: | Factor | You | Competitors | |--------|-----|-------------| | [Factor] | [Your advantage] | [Their limitation] | ### 🛡️ Objection Handling **FAQ Section** For each common objection (5-8): - Q: [Common concern] - A: [Reframe + Proof/Evidence] ### ⏰ Urgency & Scarcity **Limited Time Motivation** - Urgency trigger (time-based message) - Scarcity trigger (quantity-based message) - Consequence of delay (what they lose by waiting) ### 👑 Authority & Credibility **Company Expertise** - Founder story (brief credibility narrative) - Credentials (achievements/certifications) - Mission (company purpose and values) ### 💰 Value Stacking **Complete Package Value** - Main Offer: [Title] - $[Value] - Bonus 1: [Title] - $[Value] - Bonus 2: [Title] - $[Value] - Total Value: $[Total] → Your Investment: $[Price] ### 🔒 Risk Reversal **Money-Back Guarantee** - Guarantee type and terms - Timeframe and conditions - Why you can guarantee this ### 🎉 Final Push **Last Chance CTA** - Urgent headline - Benefit reminder - Action CTA - Reassurance text below CTA ### 📞 Footer Trust Elements - Contact information - Policies (Privacy, Terms) - Additional credibility (trust badges) ## OUTPUT FORMAT Output as structured markdown with: - Clear section headers with emojis - Bullet points for copy elements - Tables for comparisons - Specific, actionable copy for each element ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with the Hero Section 3. **NO outro/conclusion** - End with the Footer Trust Elements 4. **Every section must have specific copy** - No placeholders or generic text 5. **Persona-specific language** throughout 6. **Conversion-optimized structure** - Each section leads to the next 7. **Mobile-first consumption** - Scannable, digestible sections ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Signup flow CRO ↗' - sendPrompt: 'Analyze the signup flow for {{companyName}} — {{description}}, optimizing the conversion path from this landing page' Secondary actions (neutral styling): - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to test variations of this landing page' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, driving traffic to this landing page' - Label: 'Messaging framework ↗' sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, expanding on the landing page copy' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Lifecycle Emails Source: https://maestrix.ai/skills/create-lifecycle-emails > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create an email sequence matching the target audience context with the Customer Value Journey framework. ## CONTEXT NEEDED - Company/product description - Primary goal (onboarding, engagement, conversion, retention) - Target audience - Email count (3-12) - Duration and frequency - Custom requirements (optional) ## EMAIL SEQUENCE FRAMEWORK ### Part 1: Sequence Overview Table | Email # | Day | Journey Stage | Subject Line (7 words) | Objective | Theme | CTA Focus | |---------|-----|---------------|------------------------|-----------|-------|-----------| ### Part 2: Email Structure Summary | Email # | Subject | Preview Text | Intro Hook | Value Focus | CTA Text | Visual Element | |---------|---------|--------------|------------|-------------|----------|----------------| ### Part 3: Complete Email Drafts For each email: - **Subject Line** (exactly 7 words) - **Preview Text** (40-50 characters) - **Email Body** (minimum 250 words) - **Call-to-Action Button** - **Visual/Design Suggestion** - **Personalization Notes** ## CONTENT REQUIREMENTS - Tips connected to product category - Specific features with how-to explanations - Real statistics and examples - Authentic testimonials - Common pitfalls and solutions - Step-by-step tutorials ## WRITING STYLE - Informative and instructional tone - Professional yet approachable - Mix of sentence lengths - Clear structure: Problem → Scenario → Causes → Solutions ## BEHAVIORAL TRIGGER MATRIX Map emails to specific user behaviors: | Trigger Event | Timing | Email Type | Goal | Subject Line Pattern | |---------------|--------|------------|------|---------------------| | Signup | Immediate | Welcome | Set expectations | "Welcome to [Brand]" | | First feature use | 2 hours | Aha moment | Reinforce value | "You just did [action]!" | | Feature discovery | 24 hours | Education | Deepen engagement | "Did you know you can..." | | Inactivity (3 days) | Day 3 | Re-engagement | Bring back | "We miss you" / "Quick question" | | Goal completion | Immediate | Celebration | Reinforce success | "Congrats! You [achievement]" | | Upgrade trigger | After X uses | Upgrade prompt | Convert to paid | "Unlock more with Pro" | | Churn risk (7 days) | Day 7 | Win-back | Prevent churn | "Is everything okay?" | ## DELIVERABILITY CHECKLIST Before sending each email: - [ ] Subject line under 50 characters - [ ] Preview text optimized (40-50 chars) - [ ] Single clear CTA - [ ] Mobile-responsive design - [ ] Unsubscribe link present - [ ] From name is recognizable - [ ] No spam trigger words - [ ] Images have alt text ## A/B TESTING FRAMEWORK | Element | Test A | Test B | Metric | Duration | |---------|--------|--------|--------|----------| | Subject line | Question | Statement | Open rate | 1 week | | CTA copy | Action verb | Benefit | Click rate | 1 week | | Send time | Morning | Afternoon | Opens | 2 weeks | | Personalization | Name | None | Clicks | 1 week | ## QUALITY RULES 1. NO opening comments 2. Exactly 7-word subject lines 3. Each email provides genuine value 4. Logical progression through lifecycle 5. Clear, action-oriented CTAs 6. Include behavioral triggers for each email ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as an onboarding destination for this email sequence' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, for prospects who haven't converted' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, supporting the email campaign themes' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to optimize email subject lines and CTAs' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Llm Optimization Source: https://maestrix.ai/skills/create-llm-optimization > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a strategy to optimize content for discovery by LLMs and AI-powered search systems. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Current website URL (optional) - Key topics/keywords (optional) ## LLM OPTIMIZATION FRAMEWORK ### 1. LLM Discovery Audit **Current Visibility Assessment** - How LLMs currently describe your brand - Accuracy of AI-generated descriptions - Competitor mentions in AI responses - Gap analysis ### 2. Entity Optimization **Brand Entity Definition** - Canonical name and variations - Category/industry classification - Key attributes to associate - Relationships to other entities **Schema Markup Recommendations** - Organization schema - Product/Service schema - FAQ schema - Article schema - Review/Rating schema ### 3. Content Structure for LLM Parsing **Page-Level Requirements** - Clear H1 with entity/topic - Structured H2/H3 hierarchy - Definition-style content blocks - Bullet/numbered lists for key points - Tables for comparative data - FAQ sections with clear Q&A format **Optimal Content Patterns** | Content Type | Structure | LLM Benefit | |--------------|-----------|-------------| | Definition blocks | "X is..." opening | Direct answer extraction | | Comparison tables | Feature vs. feature | Side-by-side analysis | | Step-by-step lists | Numbered procedures | Process explanation | | FAQ pairs | Question + Answer | Direct response matching | | Statistics | Number + context | Factual citation | ### 4. Authoritative Content Strategy **Pillar Content Recommendations** - Topic clusters to own - Definitive guides to create - Data/research to publish - Expert perspectives to feature **Citation-Worthy Assets** | Asset Type | Topic | Purpose | Format | |------------|-------|---------|--------| | Original research | | | | | Industry benchmark | | | | | Expert roundup | | | | | Comprehensive guide | | | | ### 5. Technical Optimization **Crawlability Checklist** - [ ] Clean, semantic HTML - [ ] Proper meta descriptions - [ ] Open Graph tags - [ ] Twitter cards - [ ] Sitemap optimization - [ ] robots.txt configuration **Page Speed Considerations** - Core Web Vitals targets - Mobile-first indexing ### 6. Brand Association Building **Topics to Strongly Associate With** | Topic | Current Association | Target Association | Content Needed | |-------|--------------------|--------------------|----------------| **Competitor Differentiation in AI Context** - How to be recommended over competitors - Unique value props to emphasize - Gaps in competitor AI presence ### 7. Question Optimization **Questions to Rank For** | Question | Search Intent | Current Answer Quality | Content to Create | |----------|--------------|----------------------|-------------------| Target 20+ high-value questions. ### 8. Freshness Signals **Content Update Cadence** - Evergreen content: Annual review - Statistics pages: Quarterly update - Industry news: Weekly/monthly - Product info: As-needed ### 9. Source Authority Building **Backlink Strategy for AI Credibility** - High-authority targets - Industry publications - Wikipedia/knowledge bases - Academic citations ### 10. Monitoring & Measurement **LLM Response Tracking** - Regular prompt testing - Brand mention monitoring - Accuracy verification - Competitor comparison ## QUALITY RULES 1. NO opening comments 2. Actionable technical recommendations 3. Specific content structures 4. Measurable optimization goals ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, incorporating the LLM optimization insights' Secondary actions (neutral styling): - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, complementing this LLM optimization plan' - Label: 'PR angles ↗' sendPrompt: 'Create PR angles for {{companyName}} — {{description}}, to build citations and authority for AI search' - Label: 'Article from topic ↗' sendPrompt: 'Write an article for {{companyName}} — {{description}}, optimized for both LLM and traditional search' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Marketing Copy Source: https://maestrix.ai/skills/create-marketing-copy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create high-converting marketing copy using proven persuasion frameworks. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Copy purpose (awareness, consideration, conversion) - Format needed (optional) ## MARKETING COPY FRAMEWORK ### 1. Copy Strategy Foundation **Objective**: [Specific goal] **Audience**: [Who we're speaking to] **Stage**: [Awareness/Consideration/Decision] **Action**: [Desired response] **Tone**: [Voice qualities] ### 2. Headline Bank (30 Headlines) **Problem-Focused** (5) Lead with pain point **Benefit-Focused** (5) Lead with outcome **Curiosity-Focused** (5) Create information gap **News-Focused** (5) Announce something new **How-To Focused** (5) Promise transformation **Social Proof Focused** (5) Lead with credibility ### 3. Framework-Based Copy Blocks **PAS (Problem-Agitate-Solution)** - Problem paragraph - Agitation paragraph - Solution paragraph **AIDA (Attention-Interest-Desire-Action)** - Attention hook - Interest builder - Desire creator - Action prompt **BAB (Before-After-Bridge)** - Before state - After vision - Bridge (how to get there) **4 Ps (Promise-Picture-Proof-Push)** - Promise statement - Picture of success - Proof elements - Push to action **QUEST (Qualify-Understand-Educate-Stimulate-Transition)** - Qualification - Understanding - Education - Stimulation - Transition ### 4. Copy Formats **Hero Section** - Headline - Subheadline - Body copy (2-3 sentences) - CTA button - Supporting element **Feature Section** - Section headline - Feature blocks (3-4) - Icon concept - Feature title - Feature description **Testimonial Section** - Section headline - Quote integration - Attribution format **CTA Section** - Headline - Value reinforcement - Button text - Risk reversal line ### 5. Email Copy **Subject Lines** (10 options) **Preview Text** (5 options) **Opening Line** (5 hooks) **Body Copy** (full email) **CTA** (3 variations) **P.S. Line** (3 options) ### 6. Ad Copy **Short-Form** (< 100 chars) 5 variations **Medium-Form** (100-200 chars) 5 variations **Long-Form** (200+ chars) 3 variations ### 7. Landing Page Copy **Above the Fold** - Hero headline - Value prop subhead - Social proof snippet - Primary CTA **Trust Building Section** - Credibility elements - Logos/badges - Statistics **Objection Handling Section** - FAQ format - Concern → Response **Final CTA Section** - Urgency headline - Summary of value - Action button - Guarantee ### 8. Social Media Copy **LinkedIn** (3 versions) **Twitter/X** (5 versions) **Facebook** (3 versions) **Instagram** (3 versions) ### 9. Sales Collateral **One-liner** **Elevator Pitch** (30 seconds) **Positioning Statement** **Boilerplate** (100 words) ### 10. Conversion Optimization **Power Words to Include** - Urgency words - Trust words - Value words - Emotional words **Words to Avoid** - Weak modifiers - Jargon - Passive constructions ## QUALITY RULES 1. NO opening comments 2. Benefit-led, not feature-led 3. Action-oriented language 4. Multiple variations per format ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, using this marketing copy' Secondary actions (neutral styling): - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, adapted from this copy' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, using the key messages' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, based on this copy' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Marketing Psychology Source: https://maestrix.ai/skills/create-marketing-psychology > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Apply psychological principles and cognitive biases to create more persuasive marketing strategies. ## CONTEXT NEEDED - Company name and description - Product/service type - Target audience - Primary marketing challenge - Channels being used - Conversion goals ## MARKETING PSYCHOLOGY FRAMEWORK ### 1. Cognitive Biases for Marketing **Decision-Making Biases** | Bias | Definition | Marketing Application | Example | |------|------------|----------------------|---------| | Anchoring | First information influences judgment | Show high price first, then discount | "Was $199, Now $99" | | Loss Aversion | Losses feel 2x stronger than gains | Frame as what they'll miss | "Don't miss out on..." | | Social Proof | Following others' behavior | Show popularity signals | "Join 50,000+ customers" | | Authority | Trust experts and credentials | Use expert endorsements | "Recommended by [Expert]" | | Scarcity | Rare items seem more valuable | Limited availability messaging | "Only 3 left in stock" | | Reciprocity | Obligation to return favors | Give value before asking | Free tools, guides, trials | | Commitment | Consistency with past actions | Start with small asks | Free signup → Paid upgrade | **Application Matrix** | Bias | Headlines | Email | Landing Page | Pricing | Ads | |------|-----------|-------|--------------|---------|-----| | Anchoring | | | | ✓✓✓ | ✓ | | Loss Aversion | ✓✓ | ✓✓✓ | ✓✓ | ✓ | ✓✓✓ | | Social Proof | ✓ | ✓ | ✓✓✓ | ✓✓ | ✓✓ | | Authority | ✓✓ | ✓ | ✓✓✓ | | ✓✓ | | Scarcity | ✓✓✓ | ✓✓✓ | ✓✓ | ✓ | ✓✓✓ | | Reciprocity | | ✓✓✓ | ✓✓ | | ✓ | ### 2. Persuasion Frameworks **Cialdini's 6 Principles** | Principle | Definition | Implementation | |-----------|------------|----------------| | Reciprocity | Give first, ask later | Lead magnets, free trials, free tools | | Commitment | Small yes leads to big yes | Progressive engagement, micro-commitments | | Social Proof | People follow people | Testimonials, user counts, logos | | Authority | Experts are trusted | Credentials, endorsements, certifications | | Liking | We buy from who we like | Personality, values alignment, community | | Scarcity | Limited = valuable | Time limits, quantity limits, exclusivity | **AIDA Model with Psychology** | Stage | Psychology Principle | Tactics | |-------|---------------------|---------| | Attention | Novelty, Pattern interrupt | Unexpected hooks, contrarian views | | Interest | Curiosity gap, Relevance | Open loops, personalization | | Desire | Social proof, Loss aversion | Testimonials, FOMO messaging | | Action | Scarcity, Commitment | Limited offers, clear CTAs | ### 3. Emotional Triggers **Core Emotional Drivers** | Emotion | Trigger | Copy Examples | |---------|---------|---------------| | Fear | Loss, uncertainty, risk | "Don't let competitors win" | | Greed | Gain, opportunity, ROI | "10x your results" | | Guilt | Responsibility, obligation | "Your team is counting on you" | | Pride | Status, achievement, recognition | "Join the top 1%" | | Exclusivity | Belonging to elite group | "By invitation only" | | Curiosity | Unknown, mystery, discovery | "The secret most marketers miss" | | Urgency | Time pressure, deadlines | "Offer ends tonight" | | Belonging | Community, tribe, identity | "Join 50,000 marketers like you" | **Emotional Mapping by Audience** | Audience | Primary Emotions | Avoid | |----------|------------------|-------| | C-suite executives | Pride, fear of loss, authority | Hype, desperation | | Entrepreneurs | Greed, curiosity, exclusivity | Corporate speak | | Mid-level managers | Fear, belonging, guilt | Risk-taking language | | Technical buyers | Curiosity, pride, authority | Emotional manipulation | ### 4. Pricing Psychology **Price Perception Tactics** | Tactic | Mechanism | Example | |--------|-----------|---------| | Charm pricing | Left-digit effect | $99 vs $100 | | Prestige pricing | Round numbers = quality | $500 vs $499 | | Bundle anchoring | Individual > bundle price | "$897 value, yours for $297" | | Decoy effect | Middle option looks best | 3 tiers with inflated middle | | Pennies-a-day | Small daily cost perception | "$3/day" vs "$90/month" | | Payment framing | Per user/item perception | "$5/user" vs "$500/team" | **Decoy Pricing Example** ``` Basic: $49/mo → Acceptable Pro: $99/mo → Target (decoy makes it attractive) Enterprise: $299/mo → Decoy (makes Pro look like value) ``` ### 5. Copy Psychology Patterns **Headline Formulas** | Pattern | Psychology | Example | |---------|------------|---------| | How to [outcome] | Curiosity, competence | "How to 3x your email open rates" | | [Number] ways to | Specificity, completion | "7 ways to reduce churn" | | Why [action] fails | Fear, loss aversion | "Why your ads aren't converting" | | The secret to | Curiosity, exclusivity | "The secret to viral content" | | [Outcome] in [time] | Greed, impatience | "Double revenue in 90 days" | | Stop [negative] | Pain avoidance | "Stop wasting ad spend" | **Power Words by Emotion** | Emotion | Power Words | |---------|-------------| | Urgency | Now, today, immediately, deadline, limited, hurry | | Exclusivity | Secret, insider, members-only, VIP, exclusive | | Trust | Proven, guaranteed, certified, secure, trusted | | Curiosity | Hidden, revealed, discover, unlock, breakthrough | | Value | Free, bonus, save, discount, value, deal | | Fear | Risk, danger, warning, avoid, mistake, never | ### 6. Social Proof Strategies **Types of Social Proof** | Type | Effectiveness | Best Use Case | |------|---------------|---------------| | Customer count | Medium | Broad credibility | | Logo bar | High | B2B enterprise | | Testimonials | High | Feature pages, pricing | | Case studies | Very High | Long sales cycles | | Ratings/reviews | High | Considered purchases | | Expert endorsement | Very High | New markets, authority | | Media mentions | Medium | Brand building | | Real-time activity | High | E-commerce, signups | **Social Proof Optimization** | Element | Best Practice | Example | |---------|---------------|---------| | Specificity | Use exact numbers | "47,832 customers" vs "thousands" | | Relevance | Match viewer's segment | "Used by marketing teams" | | Recency | Show fresh proof | "238 signed up this week" | | Authority | Highlight impressive clients | Recognizable logo placement | | Outcomes | Focus on results | "Increased revenue by 127%" | ### 7. Scarcity & Urgency Tactics **Legitimate Scarcity Types** | Type | Mechanism | Example | |------|-----------|---------| | Time-limited | Deadline creates urgency | "Sale ends Sunday" | | Quantity-limited | Stock creates scarcity | "Only 50 spots available" | | Access-limited | Exclusivity creates desire | "Early access for members" | | Feature-limited | Bonus creates urgency | "Free bonus with signup today" | **Implementation Guidelines** | Do | Don't | |----|-------| | Use real deadlines | Create fake urgency | | Show actual inventory | Lie about availability | | Honor stated limits | Reset timers endlessly | | Explain the "why" | Use scarcity on everything | ### 8. Trust Building Psychology **Trust Signals by Stage** | Funnel Stage | Trust Signals Needed | |--------------|---------------------| | Awareness | Brand recognition, thought leadership | | Consideration | Case studies, reviews, comparisons | | Decision | Guarantees, security badges, testimonials | | Purchase | Payment security, clear policies | | Post-purchase | Confirmation, support access, community | **Reducing Purchase Anxiety** | Anxiety | Solution | Copy Example | |---------|----------|---------------| | Will it work? | Guarantee | "30-day money-back guarantee" | | Is it safe? | Security badges | "256-bit SSL encrypted" | | Is it right for me? | Testimonials | "Perfect for teams like yours" | | What if I need help? | Support visibility | "24/7 live chat support" | | Is the price fair? | Value comparison | "10x cheaper than [alternative]" | ### 9. Behavioral Nudges **Choice Architecture** | Nudge | Application | Example | |-------|-------------|---------| | Default selection | Pre-select desired option | Annual billing pre-selected | | Simplified choice | Reduce options to 3 | 3 pricing tiers, not 5 | | Highlighted option | Visual emphasis | "Most Popular" badge | | Progress indication | Completion motivation | "You're 80% there" | | Social default | Show popular choice | "Most customers choose..." | **Micro-Commitments Ladder** ``` Step 1: Read article (low commitment) ↓ Step 2: Download free guide (email trade) ↓ Step 3: Sign up for webinar (time commitment) ↓ Step 4: Start free trial (product commitment) ↓ Step 5: Upgrade to paid (financial commitment) ``` ### 10. Psychological Segmentation **Buyer Psychology Types** | Type | Motivation | Messaging Approach | |------|------------|-------------------| | Analytical | Logic, data, proof | Features, specifications, ROI | | Driver | Results, efficiency, speed | Outcomes, time savings, bottom line | | Expressive | Innovation, creativity, trend | New, exciting, innovative | | Amiable | Security, relationships, trust | Testimonials, support, community | **Adapting Message by Type** | Element | Analytical | Driver | Expressive | Amiable | |---------|------------|--------|------------|---------| | Headline | Data-driven | Results-focused | Trendy/New | Trust-focused | | Proof | Statistics | Case studies | Social buzz | Testimonials | | CTA | "Learn more" | "Get started now" | "Join the movement" | "Talk to us" | | Risk reversal | Detailed guarantee | Quick implementation | Trending status | Personal support | ### 11. Applied Psychology Audit **Current Marketing Psychology Score** | Principle | Present? | Effectiveness | Improvement | |-----------|----------|---------------|-------------| | Social Proof | | | | | Scarcity/Urgency | | | | | Authority | | | | | Emotional Triggers | | | | | Loss Aversion | | | | | Trust Signals | | | | | Clear Value Prop | | | | **Priority Recommendations** | Recommendation | Psychology Principle | Expected Impact | |----------------|---------------------|-----------------| | | | | ## QUALITY RULES 1. NO opening comments - start with bias analysis 2. Use ethical persuasion only (no manipulation) 3. Match psychology to audience type 4. Provide specific copy examples 5. Include implementation guidance ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, applying these psychological principles' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, using these persuasion techniques' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, applying these psychological triggers' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, using these persuasion frameworks' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Marketplace Listings Source: https://maestrix.ai/skills/create-marketplace-listings > **EXECUTION RULE**: Generate the complete output yourself, then render it as an interactive HTML widget using `show_widget`. Do not return raw markdown. Use only hardcoded hex colors from the Maestrix design system. Never use CSS variables. Never make a secondary API call. Generate fully optimized marketplace listing assets across Amazon, eBay, and Walmart that maximize visibility and conversion. ## CONTEXT NEEDED - Company / brand name - Product name + category - Product features - Target audience - Price point - Primary marketplace (Amazon / eBay / Walmart) - Competitor ASINs / listings (optional) ## MARKETPLACE LISTING FRAMEWORK ### 1. Amazon Listing Optimization - Product title (200 chars): Brand + Product Line + Key Feature + Size/Quantity + Color/Variant - 5 benefit-driven bullets (500 chars each) with ALL CAPS lead, embedded keywords, structured by: primary benefit, key feature, quality/materials, use cases, risk reversal - Backend search terms (250 bytes, no duplication, no commas inside phrases) - Subject matter keywords (subject, audience, attributes) ### 2. A+ Content / Enhanced Brand Content - Module 1: Brand story banner (headline + 300-word body) - Module 2: Comparison chart (us vs generic alternative) - Module 3: Image with text overlay (headline + 100-word supporting copy) - Module 4: Technical specifications block - Module 5: FAQ module (3 Q&As — common, objection, use case) ### 3. eBay Listing Optimization - eBay title (80 chars, keyword-optimized for eBay search) - Item specifics (brand, type, material, color, size, condition, category-specific) - HTML-ready description with headers, bullets, sections ### 4. Walmart Marketplace - Product name (75 chars) - 5 key features (Walmart bullet format) - Shelf description (500 chars for category pages) - Marketing description (4000 chars long-form) ### 5. Universal Listing Assets - Image requirements checklist (main, lifestyle, infographic, scale, package, detail) - Video content brief (Hook 0-3s → Problem 3-10s → Solution 10-25s → CTA 25-30s) ## QUALITY RULES 1. NO opening fluff 2. Respect every character/byte limit 3. Keywords appear naturally, no stuffing 4. Bullets lead with benefit, not feature 5. A+ Content reads like premium editorial, not specs sheet ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract **Pattern:** F (Editorial Document) Document header card (product name + marketplace badges) → section cards per marketplace (Amazon → eBay → Walmart) with uppercase labels and 14px body → pull quote highlighting top bullet → CTA block with "ready to publish" checklist. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ### What's next actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save these marketplace listings to my {{companyName}} project" - **create_product_description** (green, primary) — sendPrompt: "Create a Shopify product description for {{companyName}} that mirrors the marketplace positioning" - **create_shopping_ads** (neutral, secondary) — sendPrompt: "Create Google Shopping ads for {{companyName}} to drive traffic alongside the marketplace listings" - **seo_metafields** (neutral, secondary) — sendPrompt: "Generate SEO metafields and alt text for {{companyName}} product pages" - **pdp_audit** (neutral, secondary) — sendPrompt: "Audit my product page for {{companyName}} for conversion best practices" --- ## Skill: Create Media Angles Source: https://maestrix.ai/skills/create-media-angles > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Generate compelling story angles that make your news media-worthy. ## CONTEXT NEEDED - Company name and description - Core news or announcement - Target publications/audience - Industry context (optional) - Trending topics to leverage (optional) ## MEDIA ANGLES FRAMEWORK ### 1. News Assessment **Core Announcement** [What's actually happening] **Newsworthiness Score** | Factor | Score (1-10) | Notes | |--------|--------------|-------| | Timeliness | | | | Impact/Scale | | | | Novelty | | | | Conflict/Tension | | | | Human Interest | | | | Trend Connection | | | ### 2. Angle Categories **Industry Impact Angles** (5) How this affects the broader industry: 1. [Angle + headline + why newsworthy] 2. ... **Trend Connection Angles** (5) Connecting to larger movements: 1. [Angle + headline + trend tie-in] 2. ... **Data/Research Angles** (5) Leading with statistics: 1. [Angle + headline + data point] 2. ... **Contrarian Angles** (5) Challenging conventional wisdom: 1. [Angle + headline + counter-narrative] 2. ... **Human Interest Angles** (5) Personal stories and journeys: 1. [Angle + headline + emotional hook] 2. ... ### 3. Top 5 Angle Deep Dives For each recommended angle: **Angle Title** **The Hook** One sentence that grabs attention **Why It's Newsworthy** What makes editors care **Target Publications** Best fits for this angle **Headline Options** (3) Different ways to frame **Key Talking Points** - Point 1 - Point 2 - Point 3 **Supporting Data** Statistics or proof points **Expert Perspective** Who should comment **Potential Interview Questions** What to prepare for ### 4. Newsjacking Opportunities **Current News Hooks** | Trending Topic | Our Connection | Angle | Timeliness | |----------------|----------------|-------|------------| | | | | | **Upcoming Opportunities** | Event/Date | Our Tie-In | Preparation Needed | |------------|------------|-------------------| | | | | ### 5. Publication-Specific Angles **Business/Finance Outlets** - Focus: Numbers, market impact - Angle adaptation: [Specific spin] - Headline: [Example] **Tech Publications** - Focus: Innovation, disruption - Angle adaptation: [Specific spin] - Headline: [Example] **Industry Trade Press** - Focus: Practitioner value - Angle adaptation: [Specific spin] - Headline: [Example] **Consumer Press** - Focus: Real-world impact - Angle adaptation: [Specific spin] - Headline: [Example] ### 6. Spokesperson Preparation **Key Messages by Angle** | Angle | Primary Message | Bridge Phrase | Proof Point | |-------|-----------------|---------------|-------------| **Tough Questions & Answers** | Question | Answer | Redirect | |----------|--------|----------| ### 7. Visual Story Elements **Photo Opportunities** - Image concept - What it illustrates - Caption **Video Concepts** - Format - Key visuals - Quote/soundbite **Infographic Data** - Key statistics - Comparison points - Timeline elements ### 8. Story Arc Development **Beginning**: [Set the scene] **Conflict**: [The problem or challenge] **Journey**: [What was done] **Resolution**: [The outcome] **Future**: [What's next] ### 9. Exclusivity Strategy **Tier 1 Exclusive Offer** - Recommended outlet - Angle to offer - Embargo terms - Backup plan if declined **Tier 2 Simultaneous** - Outlets to pitch together - Customization by outlet ### 10. Content Derivatives **From Each Angle, Create:** - Press release version - Blog post - Social media thread - Email to customers - Internal announcement ## QUALITY RULES 1. NO opening comments 2. Genuinely newsworthy angles 3. Publication-specific customization 4. Prepared talking points ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Press release ↗' - sendPrompt: 'Write a press release for {{companyName}} — {{description}}, using the strongest PR angle' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create media outreach emails for {{companyName}} — {{description}}, pitching these angles' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, based on the PR angles' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building thought leadership around these angles' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Messaging Source: https://maestrix.ai/skills/create-messaging > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Produce a complete, modular multi-page website copy framework plus a messaging system that can be reused across pages, ads, sales, and enablement. ## CONTEXT NEEDED To generate messaging, the user should provide: - Company name and description - Products or services offered - Target audience or market segment - Key benefits and features (optional) - Differentiators (optional) ## ROLE You are a senior conversion copywriter, UX writer, and brand strategist. You think in positioning, JTBD, information architecture, CRO, and SEO. Write clear, outcome-driven copy that builds trust and drives action. ## GUARDRAILS Clarity first. Benefits before features. Scannable structure. Social proof near CTAs. Accessibility. Natural SEO. Risk reversal. Localization-ready language. ## MESSAGING FRAMEWORK DELIVERABLES ### 0. Sitewide Foundations #### A. Brand Positioning - Core positioning statement (25-35 words) - Brand personality: 5 traits with do/don't guidance - Brand voice & tone: principles, sample phrases, modulation for awareness → conversion #### B. Key Messages - Primary message (core promise) - 3-5 supporting messages mapped to value pillars - Elevator pitch: 60-80 words, plus a 25-30 word short version #### C. Audience-Specific Messaging (Messaging Matrix) For 3 priority segments, provide: - Segment overview: role, context, buying triggers - Primary pain(s) → desired outcome(s) - Segment value prop (1 line) - Proof points / metrics most persuasive - Top features for this segment (3-5) - Objections & rebuttals - CTA recommendation & offer Present as both prose and a compact table. #### D. Keyword & Intent Map Primary/secondary keywords per page, search intent, supporting topics, and internal link targets. #### E. Trust Signals Logos, metrics, badges, policies to surface and where to place them. ### 1. Home Page Copy - Headline options (5) and subheads (3) - Support paragraph (80-120 words) - Pillars (3-5): label + 1-sentence benefit - How it works teaser: 3 steps - Proof strip: logos + 1 standout metric - Primary/secondary CTAs with microcopy - Final closer: recap + risk reversal + CTA ### 2. Product / How It Works - Promise headline - Three-step flow: problem → action → outcome - Feature blocks (6-10): name, benefit line, 2-3 sentence explanation, proof hook - Integrations / ecosystem - Security / reliability mini-section - CTA cluster ### 3. Solutions / Use Cases Create 3-6 persona tiles. For each: - Persona label - 1-line win - 60-90 word scenario - Relevant features (3-5) - CTA ### 4. Industries (if applicable) For each priority industry (3-5): - Headline with pain framing - Outcome paragraph - Workflows/compliance - Mini-case metric - CTA ### 5. Features (catalog) Grid or table of 10-20 features grouped by themes. Each: name → benefit → 1-2 sentence description. ### 6. Pricing - Positioning paragraph - 3-4 plans (who it's for, price/"Contact," 6-8 bullets, CTA) - "What's included" checklist - Guarantees/SLAs/support paragraph - Procurement/security FAQ links ### 7. Customers / Proof - Credibility headline - Logo wall guidance - Metrics band (3-5 numbers) - 5 testimonials (≤45 words each, role, company, outcome) - Awards/certifications/analyst notes ### 8. Case Studies - 3 abstracts (120-150 words): Context → Challenge → Approach → Result (metrics) → CTA - Titles + teaser blurbs for an index page ### 9. Security & Compliance (if relevant) - Trust headline - Pledge paragraph - Standards/attestations with one-line explanations - Data handling bullets - Responsible AI statement - Whitepaper CTA ### 10. About - Mission headline - Origin story (120-160 words) - Values (3-5 with one-liners) - Leadership mini-bios - Hiring teaser ### 11. Resources / Learning - Library intro - Content types (guides, webinars, templates, benchmarks) - Newsletter signup with benefit bullets - SEO topic clusters ### 12. FAQ 10-15 Q&As covering: - Fit - Onboarding - Data/security - Pricing - Cancellation - ROI - Support - Integrations - Procurement - Compliance Link to deeper pages. ### 13. Footer - Mini value prop + CTA - Nav schema - Contact details - Legal links ## OUTPUT FORMAT Format everything in clear, actionable markdown with proper headings and structure. Focus on conversion-oriented copy that builds trust and drives action. ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with the first section 3. **NO outro/conclusion** - End with the last piece of useful content 4. **NO title duplication** - Don't repeat the task name as a heading 5. **Clarity first** - Every sentence should be immediately understandable 6. **Benefits before features** - Lead with outcomes 7. **Scannable structure** - Headers, bullets, short paragraphs 8. **Social proof near CTAs** - Reinforce action with trust ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page copy ↗' - sendPrompt: 'Create landing page copy for {{companyName}} — {{description}}, using the messaging framework we just built' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, using the messaging angles from this framework' - Label: 'Social posts ↗' sendPrompt: 'Create social media posts for {{companyName}} — {{description}}, based on this messaging framework' - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, using the objection handling and messaging from this framework' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Meta Ads Source: https://maestrix.ai/skills/create-meta-ads > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create scroll-stopping, conversion-focused Meta ad campaigns combining David Ogilvy's timeless principles with modern direct-response copywriting. ## CONTEXT NEEDED To generate Meta Ads, the user should provide: - Company name and description - Products or services offered - Target audience or persona - Campaign objective (awareness, traffic, leads, sales, etc.) - Destination URL (optional) - Additional requirements (optional) ## CAMPAIGN REQUIREMENTS Generate exactly **3 high-converting Meta ad campaigns**. For each campaign create: ### Titles (5 per campaign) - **Character limit**: 60 characters maximum (strict) - **Must include**: Relevant keywords from the project context - **Don't be generic**: Use proven copywriting frameworks ### Descriptions (5 per campaign) - **Character limit**: 90 characters maximum (strict) - **Must include**: Relevant keywords and persona-specific language - **Emphasize**: Benefits over features, emotional triggers, clear CTAs ## COPYWRITING FRAMEWORKS TO USE Mix these approaches across your variations: ❓ **Thought-Provoking Questions** - Challenge assumptions, make users think - Example: "Still manually tracking expenses in 2025?" ⚡ **Fight-the-Status-Quo Statements** - Disrupt conventional thinking - Example: "Traditional CRMs fail small teams. Here's why." 💪 **Bold Statements** - Strong claims backed by value propositions - Example: "10x your sales pipeline in 30 days, guaranteed" 🎯 **Problem-Solution Format** - Identify pain point + hint at solution - Example: "Tired of losing leads? Automate follow-ups now" ✨ **Before-After Bridge** - Imply transformation journey - Example: "From chaos to clarity: Organize your workflow" 🔥 **Urgency & Scarcity** - Time-sensitive or limited offers - Example: "48-hour flash sale: Save 50% today only" 🏆 **Social Proof Elements** - Reference authority, testimonials, results - Example: "Join 10,000+ teams who switched to [product]" 🎁 **Curiosity Gaps** - Create information gaps that demand clicks - Example: "The one feature your competitors don't want you to know" ## OUTPUT FORMAT For each of the 3 campaigns: ``` ## Campaign 1: [Campaign Name - 5 words max] **Ad Set Name:** [5 words max] **Ad Name:** [Brief identifier] ### Titles (5) 1. [Title 1 - ≤60 chars] 📊 2. [Title 2 - ≤60 chars] 🚀 3. [Title 3 - ≤60 chars] 💡 4. [Title 4 - ≤60 chars] ⚡ 5. [Title 5 - ≤60 chars] 🎯 ### Body Texts (5) 1. [Body 1 - ≤90 chars] 2. [Body 2 - ≤90 chars] 3. [Body 3 - ≤90 chars] 4. [Body 4 - ≤90 chars] 5. [Body 5 - ≤90 chars] --- ``` ## FORMATTING REQUIREMENTS - Use emojis strategically (1-2 per title/description) for visual appeal - Align copy with persona's pain points - Align copy with persona's goals - Address their primary concerns - Reference their buying triggers - Deeply leverage the company description - Reference specific products/services - Tailor messaging to the campaign objective ## CRITICAL INSTRUCTIONS 1. **Avoid Generic Copy**: Every title and description must feel custom-made 2. **Use All Context**: Reference specific details from the company/product 3. **Leverage Voice of Customer**: Mirror the language and concerns of the target audience 4. **Address Barriers Head-On**: Preemptively address objections 5. **Be Conversion-Focused**: Every word should drive toward the objective 6. **Mix Frameworks**: Vary your tactics across the 5 variations 7. **Character Limits Are Strict**: Meta will reject ads that exceed limits 8. **Persona-First Writing**: Write as if speaking directly to the target persona ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with Campaign 1 3. **NO outro/conclusion** - End with the last campaign 4. **Generate 3 campaigns total** - Each with distinct angles and messaging strategies 5. **Character limits are non-negotiable** - Count characters carefully ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Google Ads ↗' - sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, complementing the Meta Ads we just created' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a destination for these Meta Ads' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to optimize these ad campaigns' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, reinforcing the ad messaging' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create North Star Source: https://maestrix.ai/skills/create-north-star > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Define your North Star metric and build a cascading KPI framework for team alignment. ## CONTEXT NEEDED - Company name and description - Business model (B2B, B2C, marketplace, etc.) - Current stage (pre-revenue, growth, scale) - Primary revenue drivers - Current challenges or goals ## NORTH STAR FRAMEWORK ### 1. North Star Metric Definition **Primary Success Metric** | Metric Type | Definition | Why This Metric | |-------------|------------|-----------------| | Growth | Revenue, ARR, users, market share | Prioritize expansion and acquisition | | Profitability | Margins, unit economics, EBITDA | Prioritize efficiency and sustainability | | Retention | Churn rate, NRR, repeat purchases | Prioritize customer value and loyalty | | Engagement | DAU/MAU, time on platform, actions | Prioritize product stickiness | | Network Effects | Connections, referrals, virality | Prioritize ecosystem growth | **Your North Star** - Metric name: - Definition (precise calculation): - Current value: - Target value (12 months): - Why this metric matters: ### 2. North Star Validation Checklist | Criteria | Pass/Fail | Rationale | |----------|-----------|-----------| | Reflects customer value delivered | | | | Correlates with revenue growth | | | | Measurable weekly/monthly | | | | Team can influence it directly | | | | Not gameable without real progress | | | | Survives strategy pivots | | | ### 3. Supporting KPIs (Input Metrics) What activities drive the North Star? **Acquisition Metrics** | KPI | Definition | Target | Owner | |-----|------------|--------|-------| **Activation Metrics** | KPI | Definition | Target | Owner | |-----|------------|--------|-------| **Retention Metrics** | KPI | Definition | Target | Owner | |-----|------------|--------|-------| **Revenue Metrics** | KPI | Definition | Target | Owner | |-----|------------|--------|-------| **Referral Metrics** | KPI | Definition | Target | Owner | |-----|------------|--------|-------| ### 4. KPI Hierarchy Map ``` [NORTH STAR] │ ┌───────────────┼───────────────┐ │ │ │ [Level 2 KPI] [Level 2 KPI] [Level 2 KPI] │ │ │ ┌────┴────┐ ┌────┴────┐ ┌────┴────┐ [L3 KPI] [L3 KPI] [L3 KPI] [L3 KPI] [L3 KPI] [L3 KPI] ``` Fill in your specific metrics at each level. ### 5. Measurement Cadence | Metric Level | Review Frequency | Decision Trigger | Escalation Path | |--------------|------------------|------------------|-----------------| | North Star | Monthly | 10% deviation | Executive team | | Level 2 KPIs | Weekly | 15% deviation | Department head | | Level 3 KPIs | Daily | 20% deviation | Team lead | ### 6. Counter-Metrics (Guardrails) Metrics to watch so you don't over-optimize the North Star: | Counter-Metric | Minimum Threshold | Why It Matters | |----------------|-------------------|----------------| ### 7. Team Alignment Framework **Executive Level** - Primary focus: North Star + Counter-metrics - Review cadence: Monthly - Key decisions: Strategy, resource allocation **Department Level** - Primary focus: Level 2 KPIs - Review cadence: Weekly - Key decisions: Tactics, team priorities **Team Level** - Primary focus: Level 3 KPIs (activity metrics) - Review cadence: Daily/Weekly - Key decisions: Execution, experiments ### 8. Dashboard Requirements | Metric | Data Source | Visualization | Alert Threshold | |--------|-------------|---------------|-----------------| ### 9. Quarterly OKR Mapping **Objective 1: [Tied to North Star]** - KR1: - KR2: - KR3: **Objective 2: [Tied to Level 2 KPI]** - KR1: - KR2: - KR3: ### 10. North Star Communication Plan - All-hands cadence: - Dashboard access: - Celebration milestones: - Course-correction triggers: ## QUALITY RULES 1. NO opening comments 2. One clear North Star (not multiple) 3. Metrics must be measurable within current data infrastructure 4. Include ownership for every KPI 5. Counter-metrics prevent gaming ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** H **Layout spec:** Score header (large number, color-coded: >80 --success, 60-80 --warning, <60 --destructive, sub-score chips) → ranked recommendation list (severity badge, title, effort/impact chips) → quick wins (--success tint) → implementation timeline (Now/30d/90d). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'GTM strategy ↗' - sendPrompt: 'Create a GTM strategy for {{companyName}} — {{description}}, driven by this North Star metric' Secondary actions (neutral styling): - Label: 'Channel matrix ↗' sendPrompt: 'Create a channel matrix for {{companyName}} — {{description}}, optimizing for the North Star metric' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, aligned with these KPIs' - Label: 'Analytics tracking ↗' sendPrompt: 'Create an analytics tracking plan for {{companyName}} — {{description}}, measuring this North Star metric' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Offers Hooks Source: https://maestrix.ai/skills/create-offers-hooks > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Design a comprehensive incentive canvas applying behavioral economics principles from Uri Gneezy, Robert Cialdini, and proven conversion frameworks. ## CONTEXT NEEDED To generate offers and hooks, the user should provide: - Company name and description - Products or services offered - Target audience - Keywords or themes (optional) ## BEHAVIORAL ECONOMICS FRAMEWORK Apply these key principles: - **Reciprocity**: Give value first to create obligation - **Loss Aversion**: Frame offers around what they'll lose by not acting - **Scarcity**: Limited availability creates urgency - **Hyperbolic Discounting**: Immediate rewards trump future ones - **Endowment Effect**: Let them "own" something before buying - **Anchoring**: Set high reference points before revealing price - **Social Norms**: Show what others like them are doing - **Decoy Effect**: Strategic options that make the target choice obvious ## DELIVERABLES ### 1. Lead Magnet Ideas Generate 3 compelling lead magnets: - **Lead Magnet 1**: [Title + Description] - Format (guide, checklist, template, tool, etc.) - Value proposition - Target pain point addressed - **Lead Magnet 2**: [Title + Description] - **Lead Magnet 3**: [Title + Description] ### 2. Promotional Offer Concepts Generate 4 promotional offers: - **Offer 1**: [Title + Description] - Mechanism (discount, bonus, trial, guarantee) - Behavioral principle applied - Target customer stage - **Offer 2**: [Title + Description] - **Offer 3**: [Title + Description] - **Offer 4**: [Title + Description] ### 3. Attention-Grabbing Hooks Generate 5 hooks (headlines/statements): - **Hook 1**: [Headline] - **Hook 2**: [Headline] - **Hook 3**: [Headline] - **Hook 4**: [Headline] - **Hook 5**: [Headline] ### 4. Urgency/Scarcity Tactics - Recommended effective and ethical urgency tactics - Suggested scarcity elements that would work well - Implementation guidance ### 5. Value Proposition Statements - Primary value proposition - Secondary value propositions (3-4) ### 6. Behavioral Economics Incentive Canvas #### Incentive Summary Table | Incentive Name | Key Principle | Target Stage | Primary KPI | |----------------|---------------|--------------|-------------| | [Name 1] | [Principle] | [Stage] | [KPI] | | [Name 2] | [Principle] | [Stage] | [KPI] | | [Name 3] | [Principle] | [Stage] | [KPI] | | [Name 4] | [Principle] | [Stage] | [KPI] | #### Detailed Incentive Breakdown For each incentive (4 complete incentives): **[Incentive Name 1]** **Mechanism & Timeline** - Step-by-step execution process - Trigger points and implementation timeline **Principle in Action** - Core behavioral economics concept applied - Psychological drivers and how it influences decisions **Real-World Example** - Industry-relevant case study - Results achieved **Persona & Journey Integration** - Target buyer personas and psychographic match - Customer journey stage (Awareness/Conversion/Retention) - Brand positioning alignment **Cross-Channel Strategy & Performance** - Key channel tactics (email, social, landing pages) - Sample hooks and content snippets - Primary KPIs to track (CAC, CLV, CR) - Recommended tracking tools **Implementation & ROI** - Implementation costs and resource requirements - ROI projections and scaling considerations *[Repeat for remaining 3 incentives]* ### 7. Implementation Roadmap #### Phase 1: Setup & Testing (Months 1-2) - Core incentive deployment - Initial tracking setup #### Phase 2: Optimization (Months 3-4) - Performance analysis and refinement - Cross-channel scaling ### 8. Key Success Metrics - Customer Acquisition Cost (CAC) - Customer Lifetime Value (CLV) - Conversion Rate improvements - Viral coefficient and referral rates ## OUTPUT FORMAT Format everything in markdown with: - Clear headings (# for main sections, ## for subsections) - Tables for summaries - **Bold text** for critical information - *Italics* for definitions - > Blockquotes for important insights - Bullet points for lists - Blank lines between sections for readability ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with the first section 3. **NO outro/conclusion** - End with the last piece of useful content 4. **Focus on actionable, scientifically-grounded principles** 5. **Provide specific examples** that can be immediately implemented 6. **Include measurable success metrics** for each tactic ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, featuring the best offer from this batch' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, promoting these offers' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, leading with these hooks' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, highlighting these offers' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Pipeline Playbook Source: https://maestrix.ai/skills/create-pipeline-playbook > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Design a comprehensive pipeline with stage definitions, conversion tactics, and optimization strategies. ## CONTEXT NEEDED - Company name and description - Product/service type - Average deal size - Current sales cycle length - Team structure (sales, marketing) - Current conversion rates (if known) ## PIPELINE DESIGN FRAMEWORK ### 1. Pipeline Model Selection **Choose Your Pipeline Type:** | Model | Best For | Characteristics | |-------|----------|-----------------| | **Trial/Freemium** | PLG, low ACV | Self-serve → paid conversion | | **Demo/Consultation** | Mid-market, considered purchases | Human-led discovery | | **Enterprise Sales** | High ACV, complex buying | Multi-stakeholder, long cycle | | **Waitlist/Exclusivity** | New products, limited capacity | Scarcity-driven demand | | **Partner/Referral** | Network-driven markets | Third-party introduction | **Your Primary Model:** **Secondary Model (if hybrid):** ### 2. Pipeline Stage Definitions **Stage 0: Lead Source** | Source Type | Definition | Volume Expectation | Quality Indicator | |-------------|------------|-------------------|-------------------| | Inbound Marketing | | | | | Outbound Prospecting | | | | | Product Signups | | | | | Referrals | | | | | Partners | | | | **Stage 1: [Name]** - Entry criteria: - Exit criteria: - Owner: [Marketing/Sales] - Target time in stage: - Key activities: **Stage 2: [Name]** - Entry criteria: - Exit criteria: - Owner: [Marketing/Sales] - Target time in stage: - Key activities: **Stage 3: [Name]** - Entry criteria: - Exit criteria: - Owner: [Sales] - Target time in stage: - Key activities: **Stage 4: [Name]** - Entry criteria: - Exit criteria: - Owner: [Sales] - Target time in stage: - Key activities: **Stage 5: Closed-Won** - Entry criteria: - Handoff process: - Success metrics: **Closed-Lost** - Required fields: - Recycling rules: - Win-back timing: ### 3. Lead Qualification Framework **BANT+ Scoring Model** | Criteria | Weight | Score 1 | Score 3 | Score 5 | |----------|--------|---------|---------|---------| | **Budget** | 25% | No budget | Budget discussions | Approved budget | | **Authority** | 25% | Influencer only | Champions involved | Decision maker engaged | | **Need** | 20% | Nice-to-have | Acknowledged problem | Urgent priority | | **Timeline** | 15% | 12+ months | 6-12 months | <6 months | | **Fit** | 15% | Outside ICP | Partial match | Ideal ICP | **Qualification Thresholds** - MQL: Score ≥ 50 - SQL: Score ≥ 70 - Commit: Score ≥ 85 ### 4. Conversion Tactics by Stage #### Stage 1 → Stage 2 **Primary Tactics** | Tactic | Description | Success Rate | Resources | |--------|-------------|--------------|-----------| **Automation Opportunities** - Email sequences: - Scoring triggers: - Routing rules: #### Stage 2 → Stage 3 **Primary Tactics** | Tactic | Description | Success Rate | Resources | |--------|-------------|--------------|-----------| **Objection Handling** | Objection | Response Framework | Supporting Content | |-----------|-------------------|-------------------| #### Stage 3 → Stage 4 **Primary Tactics** | Tactic | Description | Success Rate | Resources | |--------|-------------|--------------|-----------| **Stakeholder Mapping** | Role | Concerns | Messaging | Champion Strategy | |------|----------|-----------|-------------------| #### Stage 4 → Closed-Won **Primary Tactics** | Tactic | Description | Success Rate | Resources | |--------|-------------|--------------|-----------| **Closing Accelerators** - Limited-time offers: - Implementation incentives: - Executive involvement triggers: ### 5. Pipeline Velocity Optimization **Velocity Formula** ``` Pipeline Velocity = (# Opportunities × Win Rate × Avg Deal Size) / Sales Cycle Length ``` **Current State** | Metric | Value | Benchmark | Gap | |--------|-------|-----------|-----| | Opportunities/month | | | | | Win rate | | | | | Avg deal size | | | | | Sales cycle (days) | | | | | **Velocity** | | | | **Improvement Levers** | Lever | Current | Target | Tactics to Improve | |-------|---------|--------|-------------------| | Volume | | +X% | | | Win Rate | | +X% | | | Deal Size | | +$X | | | Cycle Time | | -X days | | ### 6. Pipeline Health Metrics **Leading Indicators** | Metric | Target | Alert Threshold | Action Plan | |--------|--------|-----------------|-------------| | New pipeline created | | | | | Stage progression rate | | | | | Activity levels (calls, emails) | | | | | Deal engagement score | | | | **Lagging Indicators** | Metric | Target | Review Cadence | |--------|--------|----------------| | Closed-won revenue | | Weekly | | Win rate | | Monthly | | Average deal size | | Monthly | | Sales cycle length | | Monthly | ### 7. Handoff Protocols **Marketing → Sales Handoff (MQL → SQL)** - Notification method: - SLA for response: - Required information: - Feedback loop: **SDR → AE Handoff** - Qualification checklist: - Meeting scheduling: - Context transfer: - Accountability: **Sales → Customer Success Handoff** - Trigger: - Documentation: - Intro call format: - Success criteria: ### 8. Deal Inspection Framework **Weekly Pipeline Review Agenda** 1. New opportunities added 2. Stage progressions (and stalls) 3. Deals at risk (flagged) 4. Commit vs. best-case 5. Key next steps **Deal Qualification Questions** - Why will they buy? - Why will they buy from us? - Why will they buy now? - What could stop the deal? - Who else is involved? ### 9. Lost Deal Analysis **Required Capture Fields** | Field | Options | |-------|---------| | Primary loss reason | Competitor, price, timing, no decision, other | | Competitor (if applicable) | List main competitors | | Stage lost at | Which stage | | Deal value | Amount | | Learnings | Free text | **Monthly Win/Loss Review** - Win rate by segment: - Win rate by competitor: - Common objections: - Product gaps identified: ### 10. Pipeline Coverage Model **Coverage Ratio Targets** | Deal Stage | Months Out | Coverage Ratio | |------------|------------|----------------| | Commit | Current month | 1.0x | | Best Case | Current month | 1.5x | | Pipeline | Current quarter | 3-4x | | Early Stage | Next quarter | 4-5x | **Gap Analysis** | Period | Target | Current Pipeline | Coverage | Gap | |--------|--------|------------------|----------|-----| ## QUALITY RULES 1. NO opening comments 2. All stages must have entry AND exit criteria 3. Include specific conversion tactics, not generic advice 4. Velocity formula must be calculated 5. Handoff protocols must specify SLAs ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Sales script ↗' - sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, for each pipeline stage' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, for the outbound stage' - Label: 'Lifecycle emails ↗' sendPrompt: 'Create lifecycle emails for {{companyName}} — {{description}}, nurturing pipeline leads' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, supporting pipeline conversion' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Positioning Source: https://maestrix.ai/skills/create-positioning > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a complete product positioning canvas that delivers detailed, actionable insights for strategic planning. ## CONTEXT NEEDED To generate positioning, the user should provide: - Company name and description - Products or services offered - Target audience or market segment - Known competitors (optional) ## POSITIONING FRAMEWORK Generate a comprehensive positioning strategy with the following structured sections: ### 1. Market Category & Frame of Reference Define the market category frame of reference by researching credible platforms: - **Primary Market Category**: Main category and subcategories - **Research Sources**: Gartner, G2 Crowd, TrustPilot, Forrester, Amazon, industry reports - **Complete Competitor List**: ALL competitors including: - Direct competitors (same category) - Startups disrupting the space - Legacy companies with similar solutions - Repurposed tools serving similar needs - **Specific Examples**: With credible source citations ### 2. Market Statistics Summary Create a market statistics table with REAL, VERIFIABLE data: | Metric | Value | Source | |--------|-------|--------| | Market Size (TAM) | $X billion | [Credible source] | | Market Size (SAM) | $X billion | [Credible source] | | Market Size (SOM) | $X million | [Credible source] | | Growth Rate | X% CAGR | [Credible source] | | Key Market Drivers | [List] | [Credible source] | | Geographic Distribution | [Breakdown] | [Credible source] | | Customer Spending Patterns | [Insights] | [Credible source] | **CRITICAL**: Use only real, verifiable statistics from credible sources (industry reports, government data, research firms). Include source citations. NO hallucinations or made-up statistics. ### 3. Competitive Analysis Table | Competitor | Category | How They Address Market | Key Strengths | Downsides | How You're Better/Different | |------------|----------|-------------------------|---------------|-----------|---------------------------| | [Name] | [Type] | [Approach] | [Strengths] | [Weaknesses] | [Your Advantage] | Include: - Direct competitors - Indirect alternatives - Legacy solutions - DIY approaches ### 4. Unique Attributes Matrix | Unique Attribute | Detailed Description | Why It's Unique | Customer Pain Point Solved | Business Impact | |------------------|---------------------|-----------------|---------------------------|-----------------| | [Attribute] | [Description] | [Uniqueness] | [Pain Point] | [Impact] | **REQUIREMENT**: Focus on attributes that ONLY this company possesses, not shared industry features. ### 5. Value Proposition & Proof #### Core Value Delivered - Primary benefits and outcomes - Jobs-to-be-done enabled - Quantifiable business impact #### Proof Points (with credible sources) - Case studies and success stories - Customer testimonials and quotes - ROI metrics and performance data - Certifications and awards - Third-party validation **REQUIREMENT**: Include links to credible sources where possible. Support all claims with evidence. ### 6. Customer Segments Analysis | Segment | Demographics | Psychographics | Core Motivations | Buying Triggers | Barriers to Purchase | Value Perception | |---------|--------------|----------------|------------------|-----------------|---------------------|------------------| | [Name] | [Age, size, industry] | [Values, attitudes, lifestyle] | [What drives them] | [When they buy] | [What stops them] | [How they see value] | **REQUIREMENT**: Include detailed psychographic insights (values, attitudes, lifestyle factors) alongside demographics. ### 7. Positioning Statement Framework **For** [target customer] **Who** [customer need/opportunity] **[Company] is** [market category] **That** [key benefit/differentiation] **Unlike** [primary alternative] **We** [unique advantage/proof] ### 8. Messaging Hierarchy #### Primary Message [Core positioning message] #### Supporting Messages 1. [Key differentiator message] 2. [Value proposition message] 3. [Proof/credibility message] #### Proof Points - [Evidence 1] - [Evidence 2] - [Evidence 3] ### 9. SWOT Analysis #### Strengths - [Internal advantages] - [Unique capabilities] - [Market position strengths] #### Weaknesses - [Internal limitations] - [Areas for improvement] - [Competitive gaps] #### Opportunities - [Market opportunities] - [Emerging trends] - [Expansion possibilities] #### Threats - [Competitive threats] - [Market risks] - [External challenges] ### 10. Strategic Recommendations #### Immediate Actions (0-3 months) - [Priority positioning initiatives] #### Short-term Strategy (3-6 months) - [Market positioning moves] #### Long-term Vision (6-12 months) - [Strategic positioning goals] ## OUTPUT FORMAT Output as structured markdown with clear headings, tables, and bullet points. ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with the first section 3. **NO outro/conclusion** - End with the last piece of useful content 4. **Use ONLY credible, verifiable sources** - NO HALLUCINATIONS 5. **Provide specific examples and citations** with links where possible 6. **Focus on actionable insights** with precise, concise descriptions 7. **Ensure all statistics are real** and sourced from credible platforms 8. **Support ALL claims with evidence** 9. **Focus on attributes that ONLY this company has** 10. **Include detailed psychographic analysis** in customer segments ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, building on the positioning we just defined' Secondary actions (neutral styling): - Label: 'Landing page copy ↗' sendPrompt: 'Create landing page copy for {{companyName}} — {{description}}, using this positioning strategy' - Label: 'Battle cards ↗' sendPrompt: 'Analyze the top competitor of {{companyName}} — {{description}}, using the competitive positioning from this analysis' - Label: 'Brand voice ↗' sendPrompt: 'Define the brand voice for {{companyName}} — {{description}}, aligned with this positioning' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Press Release Source: https://maestrix.ai/skills/create-press-release > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create newsworthy press releases following AP style guidelines with distribution strategy. ## CONTEXT NEEDED - Company name and description - Announcement type (product, funding, partnership, etc.) - Key details of announcement - Spokesperson name and title - Relevant metrics/proof points (optional) ## PRESS RELEASE FRAMEWORK ### 1. News Angle Assessment **Newsworthiness Factors** | Factor | Strength (1-10) | How to Emphasize | |--------|-----------------|------------------| | Timeliness | | | | Impact | | | | Prominence | | | | Proximity | | | | Novelty | | | | Human interest | | | **Best Angle**: [Recommended lead] ### 2. Press Release Structure **HEADLINE** [Action verb + key news + benefit/impact] Character limit: 80-100 **SUBHEAD** [Supporting detail that adds context] Character limit: 120-150 **DATELINE** [CITY, State – Date] – **LEAD PARAGRAPH** (Who, What, When, Where, Why) [The most newsworthy information in 2-3 sentences. Answer: What happened and why it matters.] **SECOND PARAGRAPH** (Expansion) [Key details, context, and significance.] **QUOTE 1** (Executive/Spokesperson) "[Statement that adds perspective, vision, or emotional element not in the lead.]" **BODY PARAGRAPHS** (3-4 paragraphs) [Supporting details in order of importance] - Background and context - How it works/details - Market impact - Customer/partner perspective **QUOTE 2** (Customer/Partner/Analyst - optional) "[Third-party validation quote]" **BOILERPLATE** [Standard company description - 50-75 words] **CONTACT INFORMATION** Media Contact: [Name] [Title] [Email] [Phone] ### 3. Headline Options (5) 1. [News-focused] 2. [Benefit-focused] 3. [Impact-focused] 4. [Trend-focused] 5. [Data-focused] ### 4. Quote Bank **Executive Quotes** (3 options) Focus: Vision, strategy, commitment **Customer Quotes** (2 options) Focus: Value, results, recommendation **Partner Quotes** (2 options) Focus: Collaboration, joint value ### 5. Key Messages **Primary Message** [One sentence core takeaway] **Supporting Messages** (3) 1. [Message + proof point] 2. [Message + proof point] 3. [Message + proof point] ### 6. SEO Optimization **Target Keywords** - Primary: [keyword] - Secondary: [2-3 keywords] **Optimized Elements** - Headline with keyword - First paragraph keyword - Anchor text opportunities ### 7. Multimedia Assets **Suggested Assets** - High-res logo - Executive headshot - Product image/screenshot - Infographic - Video embed **Asset Descriptions** Caption text for each visual ### 8. Distribution Strategy **Wire Services** - Tier 1: [BusinessWire, PR Newswire] - Tier 2: [Regional services] **Direct Outreach List** | Publication | Reporter | Beat | Why Relevant | |-------------|----------|------|--------------| **Embargo Strategy** - Embargo date/time - Exclusive offer consideration ### 9. Supporting Materials **Fact Sheet** Key points in bullet format **FAQ** Anticipated questions and answers **Background** Extended company/product information ### 10. Announcement Variations **Short Version** (150 words) For social, email teasers **Medium Version** (300 words) For blog, newsletter **Long Version** (500+ words) Full press release **Tweet Thread** Breaking down announcement **LinkedIn Post** Executive voice version ## QUALITY RULES 1. NO opening comments 2. AP style formatting 3. Inverted pyramid structure 4. Quotes that add, not repeat 5. Newsworthy angle emphasized ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Media angles ↗' - sendPrompt: 'Create PR pitch angles for {{companyName}} — {{description}}, building on this press release' Secondary actions (neutral styling): - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, announcing this news' - Label: 'Cold email sequence ↗' sendPrompt: 'Create outreach emails for {{companyName}} — {{description}}, pitching this story to journalists' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, maximizing the PR momentum' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Pricing Strategy Source: https://maestrix.ai/skills/create-pricing-strategy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive pricing strategy with tier design, psychological pricing, and competitive positioning. ## CONTEXT NEEDED - Company name and description - Product/service type - Current pricing (if any) - Target customer segments - Key competitors and their pricing - Primary value metrics - Cost structure overview ## PRICING STRATEGY FRAMEWORK ### 1. Value Metric Analysis **Identifying the Right Value Metric** | Potential Metric | Scales with Value? | Easy to Understand? | Predictable? | Recommendation | |------------------|-------------------|---------------------|--------------|----------------| | Per user/seat | | | | | | Per usage/volume | | | | | | Per feature tier | | | | | | Flat rate | | | | | | Hybrid | | | | | **Value Metric Selection** ``` Recommended Value Metric: [Metric] Rationale: [Why this metric aligns with customer value] Implementation: [How customers will be charged] ``` **Usage Patterns Analysis** | Segment | Typical Usage | Value Received | Willingness to Pay | |---------|---------------|----------------|-------------------| | Small/Starter | | | $ | | Mid-market | | | $ | | Enterprise | | | $ | ### 2. Competitive Pricing Analysis **Competitive Landscape** | Competitor | Entry Price | Mid-tier | Enterprise | Value Metric | Positioning | |------------|-------------|----------|------------|--------------|-------------| | | $ | $ | $ | | | **Price Positioning Map** ``` HIGH PRICE ↑ [Competitor A] [Your position?] │ LOW ←──────┼──────→ HIGH FEATURES │ FEATURES │ [Competitor B] [Competitor C] ↓ LOW PRICE ``` **Competitive Positioning Recommendation** - [ ] Premium (Above market) - [ ] Market (Match competitors) - [ ] Penetration (Below market) - [ ] Value (More for same) ### 3. Tier Structure Design **Tier Architecture** ``` [FREE/TRIAL] → [STARTER] → [PROFESSIONAL] → [ENTERPRISE] ↓ ↓ ↓ ↓ [Purpose] [Purpose] [Purpose] [Purpose] [Target] [Target] [Target] [Target] [Limits] [Price] [Price] [Custom] ``` **Detailed Tier Specification** | Feature/Limit | Free | Starter | Professional | Enterprise | |---------------|------|---------|--------------|------------| | Price (Monthly) | $0 | $ | $ | Custom | | Price (Annual) | $0 | $ | $ | Custom | | [Core feature 1] | ✓/Limit | ✓ | ✓ | ✓ | | [Core feature 2] | ✗ | ✓ | ✓ | ✓ | | [Advanced feature 1] | ✗ | ✗ | ✓ | ✓ | | [Advanced feature 2] | ✗ | ✗ | ✗ | ✓ | | [Support level] | Community | Email | Priority | Dedicated | | [Usage limit 1] | X | X | X | Unlimited | | [Usage limit 2] | X | X | X | Unlimited | ### 4. Psychological Pricing Tactics **Anchoring Strategy** ``` Display Order: [Enterprise] → [Professional] → [Starter] → [Free] Default/Highlighted: [Professional - "Most Popular"] Anchor Effect: Enterprise price makes Professional seem reasonable ``` **Price Point Psychology** | Current Price | Optimized Price | Psychology | |---------------|-----------------|------------| | $100/mo | $99/mo | Charm pricing (left digit effect) | | $99/mo | $97/mo | Odd pricing (perceived discount) | | $500/mo | $497/mo | Premium charm pricing | **Framing Tactics** | Tactic | Implementation | Example | |--------|----------------|---------| | Daily cost | Annual ÷ 365 | "Less than $1/day" | | Per user/item | Total ÷ units | "$10/user/month" | | ROI framing | Value ÷ cost | "10x ROI within 90 days" | | Comparison anchor | vs. alternative | "1/10th the cost of [X]" | | Time saved | Hours × hourly rate | "Saves 10 hrs/week = $2,400/mo" | ### 5. Freemium/Trial Strategy (if applicable) **Free Tier Design** | Element | Recommendation | Rationale | |---------|----------------|-----------| | Features included | [List] | Enough to show value | | Usage limits | [Limits] | Triggers upgrade need | | Time limit | None / X days | | | Upgrade triggers | [Events] | Natural expansion points | | Conversion target | X% | Industry benchmark | **Trial Strategy** | Option | Duration | Best For | |--------|----------|----------| | Free trial (no card) | 14 days | High-touch products | | Free trial (card required) | 7-14 days | Higher conversion intent | | Reverse trial (start premium) | 14 days | Feature discovery | | Sandbox/demo | Unlimited | Complex products | **Upgrade Trigger Points** | Trigger | Message | Urgency | |---------|---------|---------| | Usage limit hit | "You've reached your limit" | High | | Feature gated | "Available on Pro" | Medium | | Team growth | "Add team members" | Medium | | Success moment | "Unlock more with Pro" | Low | ### 6. Pricing Page Optimization **Page Structure** ``` [Social Proof / Trust Bar] ↓ [Value Proposition Headline] ↓ [Toggle: Monthly / Annual (Savings Badge)] ↓ [Tier Cards - 3-4 options] ↓ [Feature Comparison Table] ↓ [FAQ Section] ↓ [Enterprise CTA] ``` **Tier Card Best Practices** - Highlight recommended tier with "Most Popular" badge - Use contrasting color for primary CTA - Show annual savings prominently (e.g., "Save 20%") - Include key differentiators, not all features - Place social proof within cards **Messaging for Each Tier** | Tier | Tagline | Target Persona | Key Differentiator | |------|---------|----------------|-------------------| | Free | | | | | Starter | | | | | Professional | | | | | Enterprise | | | | ### 7. Discount Strategy **Discount Framework** | Discount Type | Amount | When to Use | Risks | |---------------|--------|-------------|-------| | Annual prepay | 15-20% | Default offer | Cash flow dependency | | Volume/seats | 10-30% | Expansion | Margin compression | | Startup/nonprofit | 25-50% | Market expansion | Expectation anchoring | | Competitive switch | 20-30% | Displacement | Price war | | Win-back | 25-40% | Churn reduction | Sets precedent | **Discount Governance** - Maximum discount authority by role - Required documentation for exceptions - Sunset clauses for promotional pricing - Grandfathering policy for price increases ### 8. Expansion Revenue Strategy **Upsell Pathways** | From → To | Trigger | Message | Target Rate | |-----------|---------|---------|-------------| | Free → Starter | [Event] | [Message] | % | | Starter → Pro | [Event] | [Message] | % | | Pro → Enterprise | [Event] | [Message] | % | **Cross-sell Opportunities** | Add-on | Price | Trigger | Attach Rate Target | |--------|-------|---------|-------------------| | | $ | | % | **Net Revenue Retention Target** - Target NRR: 110%+ - Expansion drivers: [List] - Contraction prevention: [Strategies] ### 9. Price Testing Roadmap **A/B Tests to Run** | Test | Hypothesis | Metric | Priority | |------|------------|--------|----------| | Price points | $X vs $Y conversion | Conversion rate | High | | Tier naming | Professional vs Business | Perception | Medium | | Annual discount | 15% vs 20% | Annual adoption | Medium | | Default tier | Starter vs Pro | ARPU | High | | Feature packaging | Bundle A vs B | Upgrade rate | Medium | **Van Westendorp Analysis Questions** 1. At what price would [product] be too expensive to consider? 2. At what price would [product] seem expensive but still worth it? 3. At what price would [product] be a bargain? 4. At what price would [product] be so cheap you'd question quality? ### 10. Implementation Checklist **Launch Readiness** - [ ] Pricing page design finalized - [ ] Billing system configured - [ ] Feature gates implemented - [ ] Upgrade/downgrade flows tested - [ ] Email templates for billing events - [ ] FAQ and support documentation - [ ] Sales team trained on positioning - [ ] Competitive battlecards updated **Metrics to Track** | Metric | Definition | Target | |--------|------------|--------| | Conversion rate | Trials → Paid | % | | ARPU | Avg revenue per user | $ | | NRR | Net revenue retention | % | | Tier distribution | % in each tier | | | Annual adoption | % paying annually | % | ## QUALITY RULES 1. NO opening comments - start with value metric analysis 2. Include specific price point recommendations 3. Use psychological pricing principles 4. Design for clear upgrade pathways 5. Include competitive positioning context ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C **Layout spec:** Header row (one card per competitor with score badge) → feature matrix (color-coded: --success win, --destructive lose, --muted neutral) → battle card block (positioning, weakness, counter) → bottom highlights (--warning gap, --success opportunity). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a pricing landing page for {{companyName}} — {{description}}, showcasing the pricing strategy we just defined' Secondary actions (neutral styling): - Label: 'Value proposition ↗' sendPrompt: 'Refine the value proposition for {{companyName}} — {{description}}, aligned with this pricing strategy' - Label: 'Messaging framework ↗' sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, incorporating price-value positioning' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, highlighting the pricing advantage' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Product Description Source: https://maestrix.ai/skills/create-product-description > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create conversion-optimized product descriptions that highlight benefits and drive purchases. ## CONTEXT NEEDED - Company name and description - Product name and details - Target customer - Key features/specifications - Price point (optional) - Competitor products (optional) ## PRODUCT DESCRIPTION FRAMEWORK ### 1. Product Positioning **Category**: [Product category] **Price Position**: [Budget/Mid/Premium] **Primary Benefit**: [One sentence] **Key Differentiator**: [What sets it apart] **Target Buyer**: [Specific persona] ### 2. Description Lengths **Ultra-Short** (15 words) For ads, search results, quick reference **Short** (50 words) For category pages, comparison tables **Medium** (150 words) For product cards, email features **Full** (300-500 words) For product detail pages **Extended** (750+ words) For SEO-rich landing pages ### 3. Full Product Description Structure **Opening Hook** (1-2 sentences) Problem-agitation or benefit-first lead **Key Benefit Statement** (2-3 sentences) Primary value proposition **Feature-Benefit Blocks** (5-7) | Feature | Benefit | Proof | |---------|---------|-------| **Use Case Scenarios** (3-5) "Perfect for..." situations **Social Proof Integration** - Customer quote - Rating summary - Usage statistic **Technical Specifications** - Spec table - Compatibility info - Dimensions/materials **Objection Handling** Preemptively address concerns **Guarantee/Risk Reversal** Returns, warranty, satisfaction promise **Call to Action** Urgency elements, buying prompt ### 4. SEO Optimization **Primary Keyword**: [Target term] **Secondary Keywords**: [3-5 related terms] **Long-tail Phrases**: [5-10 specific queries] **Optimized Title Tag** (60 chars) **Meta Description** (160 chars) ### 5. Emotional Triggers **Pain Points Addressed** - Pain 1 → How product solves - Pain 2 → How product solves **Aspirational Benefits** - What customer becomes - Status/identity shift **Sensory Language** Words that evoke experience ### 6. Trust Elements - Manufacturing quality - Certifications - Awards - Media mentions - Expert endorsements ### 7. Comparison Copy **vs. Alternative A** | Aspect | Our Product | Alternative | |--------|-------------|-------------| **vs. Alternative B** Why customers switch to us ### 8. Channel Adaptations **E-commerce Listing** (Amazon, Shopify) - Title optimization - Bullet points (5) - A+ content concepts **Social Commerce** - Instagram Shop caption - Facebook Shop description **Email Feature** - Subject line - Preview text - Body copy ### 9. Seasonal/Promotional Variants **Holiday Version** Gift-giving angle **Sale Version** Urgency and value emphasis **Bundle Version** Package value proposition ### 10. Testing Recommendations **Elements to A/B Test** - Headlines - Benefit order - CTA wording - Price presentation - Social proof placement ## QUALITY RULES 1. NO opening comments 2. Benefits before features 3. Scannable format with bullets 4. SEO keywords naturally integrated ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, expanding on this product description' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, using messaging from this product description' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, highlighting key product features' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, introducing this product' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Product Insights Source: https://maestrix.ai/skills/create-product-insights > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create actionable product insights from customer feedback, market data, and competitive analysis. ## CONTEXT NEEDED - Company name and description - Product or feature focus - Customer feedback sources (optional) - Competitive context (optional) - Key metrics to improve (optional) ## PRODUCT INSIGHTS FRAMEWORK ### 1. Product-Market Fit Assessment **Current PMF Score Indicators** - Customer retention rate - NPS/satisfaction scores - Organic growth rate - Usage frequency - Expansion revenue **PMF Gap Analysis** | Dimension | Current State | Ideal State | Gap | Priority | |-----------|---------------|-------------|-----|----------| ### 2. Voice of Customer Synthesis **Praise Themes** (What customers love) | Theme | Frequency | Customer Quotes | Implication | |-------|-----------|-----------------|-------------| **Pain Point Themes** (What frustrates customers) | Theme | Frequency | Customer Quotes | Implication | |-------|-----------|-----------------|-------------| **Request Themes** (What customers want) | Theme | Frequency | Customer Quotes | Business Impact | |-------|-----------|-----------------|-----------------| ### 3. Usage Pattern Insights **Power User Behaviors** - Actions that correlate with retention - Features used by best customers - Workflows that drive value **Churn Indicators** - Early warning signals - Features NOT used by churned customers - Time-to-value gaps ### 4. Competitive Feature Analysis | Feature | Our Product | Competitor A | Competitor B | Gap/Opportunity | |---------|-------------|--------------|--------------|-----------------| ### 5. Feature Prioritization Matrix | Feature/Improvement | Customer Impact | Business Impact | Effort | Priority Score | |---------------------|-----------------|-----------------|--------|----------------| **Scoring Methodology** - Customer Impact: 1-5 (reach × satisfaction improvement) - Business Impact: 1-5 (revenue/retention effect) - Effort: 1-5 (inverse, lower = easier) - Priority = (Customer × Business) / Effort ### 6. Value Proposition Validation **Current Value Props vs. Customer Reality** | Claimed Benefit | Customer Agreement | Evidence | Adjustment Needed | |-----------------|-------------------|----------|-------------------| ### 7. Segmentation Insights | Segment | Unique Needs | Underserved Areas | Opportunity Size | |---------|--------------|-------------------|------------------| ### 8. Messaging Opportunities Based on insights, recommended messaging updates: - Features to emphasize more - Benefits to reframe - Proof points to highlight - Objections to address proactively ### 9. Roadmap Recommendations **Quick Wins** (< 1 month) - Item, impact, rationale **Strategic Bets** (1-3 months) - Item, impact, rationale **Foundation Work** (3-6 months) - Item, impact, rationale ### 10. Metrics to Track | Insight Area | Metric | Current | Target | Timeline | |--------------|--------|---------|--------|----------| ## QUALITY RULES 1. NO opening comments 2. Data-driven recommendations 3. Specific, actionable insights 4. Clear prioritization logic ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** H **Layout spec:** Score header (large number, color-coded: >80 --success, 60-80 --warning, <60 --destructive, sub-score chips) → ranked recommendation list (severity badge, title, effort/impact chips) → quick wins (--success tint) → implementation timeline (Now/30d/90d). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Positioning ↗' - sendPrompt: 'Create positioning for {{companyName}} — {{description}}, building on these product insights' Secondary actions (neutral styling): - Label: 'Persona ↗' sendPrompt: 'Generate a persona for {{companyName}} — {{description}}, based on the use cases identified' - Label: 'Messaging framework ↗' sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, organized by these use cases' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, showcasing these use cases' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Programmatic Seo Source: https://maestrix.ai/skills/create-programmatic-seo > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a programmatic SEO strategy with page templates, data structures, and implementation roadmaps for scalable organic traffic. ## CONTEXT NEEDED - Company name and description - Products/services offered - Target keywords and topics - Available data sources - Technical implementation capacity - Content volume goals ## PROGRAMMATIC SEO FRAMEWORK ### 1. pSEO Opportunity Analysis **Template Opportunity Identification** | Template Type | URL Pattern | Example | Est. Pages | Traffic Potential | |---------------|-------------|---------|------------|-------------------| | Location | /[service]-in-[city] | /marketing-agency-in-austin | | | | Integration | /integrations/[tool] | /integrations/salesforce | | | | Comparison | /[brand]-vs-[competitor] | /hubspot-vs-salesforce | | | | Use case | /[use-case]-software | /email-marketing-software | | | | Industry | /[industry]-solutions | /healthcare-solutions | | | | Template | /[type]-templates | /invoice-templates | | | | Calculator | /[type]-calculator | /roi-calculator | | | | Directory | /[category]/[item] | /tools/crm | | | **Keyword Pattern Research** | Pattern | Modifier Examples | Search Volume Range | Competition | |---------|-------------------|---------------------|-------------| | [service] in [city] | 500+ cities | | | | [brand] alternative | 50+ competitors | | | | [tool] integration | 100+ tools | | | | best [category] for [use case] | 20+ use cases | | | ### 2. Data Source Mapping **Internal Data Sources** | Data Type | Source | Fields Available | Update Frequency | |-----------|--------|------------------|------------------| | Products | Database | name, description, price, specs | Real-time | | Locations | API | city, state, zip, coordinates | Static | | Integrations | CMS | name, logo, features, status | Weekly | | Categories | Taxonomy | name, description, parent | Monthly | **External Data Sources** | Data Type | Source | Access Method | Cost | |-----------|--------|---------------|------| | City data | Census/Wikipedia | API/Scrape | Free | | Company info | Clearbit/LinkedIn | API | $$ | | Keywords | Ahrefs/SEMrush | Export | $$ | | Reviews | G2/Capterra | API | $-$$ | **Data Quality Requirements** | Field | Required? | Validation | Fallback | |-------|-----------|------------|----------| | Unique title | Yes | Not duplicate | Generate from pattern | | Meta description | Yes | 120-160 chars | Template + variables | | H1 | Yes | Contains keyword | Same as title | | Body content | Yes | Min 500 words | Template + data fill | | Images | Recommended | Alt text present | Stock/placeholder | ### 3. Page Template Design **Template Structure** ``` [Header / Navigation] ↓ [H1: Dynamic Title] ↓ [Hero: Key Value + CTA] ↓ [Section 1: Overview / Intro - 100-200 words] ↓ [Section 2: Key Features / Benefits - 200-300 words] ↓ [Section 3: Data-Driven Content (tables, stats)] ↓ [Section 4: FAQ (3-5 questions)] ↓ [Section 5: Related Items / Internal Links] ↓ [CTA Section] ↓ [Footer] ``` **SEO Element Templates** **Title Tag Template** ``` Primary: [Keyword] - [Brand] | [Value Prop] Example: Email Marketing Software for E-commerce - Acme | Free Trial Variations: - [Keyword]: [Value Prop] | [Brand] - Best [Keyword] in 2024 | [Brand] - [Keyword] - [Modifier] | [Brand] ``` **Meta Description Template** ``` Primary: [Hook/Question] [Solution with keyword]. [Proof point]. [CTA]. Example: Looking for email marketing software? Acme helps 10,000+ e-commerce brands automate campaigns. Start your free trial today. Max: 155-160 characters Variables: {keyword}, {proof_point}, {cta} ``` **H1 Template** ``` Primary: [Keyword] [Modifier] Example: Email Marketing Software for E-commerce Brands Always include primary keyword ``` ### 4. Content Generation Framework **Dynamic Content Sections** | Section | Content Type | Data Source | Uniqueness Strategy | |---------|--------------|-------------|---------------------| | Intro | Templated + variables | Keyword data | City/industry specific stats | | Features | Database-driven | Product data | Use-case framing | | Comparison | Matrix/table | Competitor data | Unique scoring | | FAQ | Pattern-based | Keyword research | Location/context specific | | Social proof | Curated | Reviews/logos | Industry-matched | **Content Uniqueness Requirements** | Page Type | Min Unique Content | Approach | |-----------|-------------------|----------| | Location pages | 40%+ | Local stats, testimonials, regulations | | Comparison pages | 60%+ | Unique scoring, detailed analysis | | Integration pages | 50%+ | Use cases, setup guides, features | | Use case pages | 70%+ | Industry-specific benefits, examples | **Avoiding Thin Content** - Minimum 500 words per page - At least 3 unique paragraphs - Include page-specific data points - Add relevant internal links - Include unique images/graphics ### 5. Internal Linking Strategy **Hub & Spoke Model** ``` [Hub: Main Topic] │ ┌─────────────────┼─────────────────┐ ↓ ↓ ↓ [Spoke 1] [Spoke 2] [Spoke 3] Category A Category B Category C │ │ │ ┌───┴───┐ ┌───┴───┐ ┌───┴───┐ ↓ ↓ ↓ ↓ ↓ ↓ [Page] [Page] [Page] [Page] [Page] [Page] ``` **Automated Internal Links** | Link Type | Placement | Anchor Text Pattern | |-----------|-----------|---------------------| | Related items | Bottom section | "[Related item name]" | | Parent category | Breadcrumb | "[Category name]" | | Sibling pages | Sidebar | "Also see: [page title]" | | Hub page | Intro paragraph | "[Topic] solutions" | ### 6. Technical Implementation **URL Structure** ``` Clean: /integrations/salesforce Avoid: /integrations.php?id=salesforce Avoid: /integrations/salesforce-integration-page ``` **Sitemap Strategy** | Sitemap | Max URLs | Update Frequency | |---------|----------|------------------| | Main sitemap | Index only | Weekly | | pSEO sitemap 1 | 50,000 | Weekly | | pSEO sitemap 2 | 50,000 | Weekly | | News sitemap | 1,000 | Daily | **Indexing Management** | Page Status | Robots | Sitemap | Google Indexing API | |-------------|--------|---------|---------------------| | Quality page | index, follow | Include | Submit | | Thin content | noindex | Exclude | Don't submit | | Duplicate | canonical to main | Exclude | Don't submit | | Seasonal | index when relevant | Include | Submit | ### 7. Quality Control Framework **Page Quality Checklist** - [ ] Unique title (not duplicate) - [ ] Meta description present and unique - [ ] H1 contains target keyword - [ ] Minimum 500 words of content - [ ] At least 3 internal links - [ ] At least 1 image with alt text - [ ] Page loads in <3 seconds - [ ] Mobile-friendly layout - [ ] No broken links - [ ] No duplicate content detected **Automated Quality Signals** | Signal | Threshold | Action if Fail | |--------|-----------|----------------| | Word count | <300 | noindex | | Duplicate title | >90% match | Regenerate | | Missing H1 | Absent | Block publish | | Broken links | Any | Queue for fix | | Load time | >5s | Optimize | ### 8. Scaling Roadmap **Phase 1: Foundation (Month 1)** - [ ] Identify top 3 template opportunities - [ ] Set up data sources - [ ] Create 1 template design - [ ] Build 50-100 test pages - [ ] Monitor indexing and traffic **Phase 2: Validation (Month 2-3)** - [ ] Analyze initial performance - [ ] Optimize template based on data - [ ] Add 2nd template type - [ ] Scale to 500-1,000 pages - [ ] Implement quality monitoring **Phase 3: Scale (Month 4-6)** - [ ] Add 3rd-4th template types - [ ] Scale to 5,000-10,000 pages - [ ] Implement automated updates - [ ] Add new data sources - [ ] Expand internal linking **Phase 4: Optimization (Ongoing)** - [ ] A/B test templates - [ ] Refresh stale content - [ ] Expand to new keywords - [ ] Improve conversion elements - [ ] Monitor for algorithm changes ### 9. Performance Measurement **Traffic Metrics** | Metric | Target | Measurement | |--------|--------|-------------| | Indexed pages | 95%+ of quality pages | GSC | | Organic traffic | [Target]/mo | GA4 | | Avg. position | Top 10 for targets | GSC | | CTR | >3% average | GSC | **Conversion Metrics** | Metric | Target | Measurement | |--------|--------|-------------| | Conversion rate | >1% | GA4 | | Pages per session | >2 | GA4 | | Bounce rate | <60% | GA4 | | Time on page | >2 min | GA4 | **Quality Metrics** | Metric | Target | Measurement | |--------|--------|-------------| | Duplicate content | <5% | Screaming Frog | | Thin pages | <10% | Custom audit | | Orphan pages | 0 | Sitemap analysis | | Load time | <3s | PageSpeed | ### 10. Common pSEO Mistakes **Avoid These Patterns** | Mistake | Why It's Bad | Fix | |---------|--------------|-----| | Same content, different URLs | Google consolidates/ignores | Unique content per page | | Auto-generated gibberish | Quality penalties | Human-quality templates | | No internal links | Poor discovery | Automated linking | | Missing local intent | Irrelevant traffic | Location-specific content | | Ignoring user intent | High bounce | Match content to intent | ## QUALITY RULES 1. NO opening comments - start with opportunity analysis 2. Include specific template examples 3. Calculate realistic traffic potential 4. Provide implementation phases 5. Include quality control mechanisms ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, incorporating programmatic SEO pages' Secondary actions (neutral styling): - Label: 'LLM optimization ↗' sendPrompt: 'Create an LLM optimization plan for {{companyName}} — {{description}}, for the programmatic pages' - Label: 'Landing page ↗' sendPrompt: 'Create a template landing page for {{companyName}} — {{description}}, for programmatic SEO' - Label: 'Analytics tracking ↗' sendPrompt: 'Create an analytics tracking plan for {{companyName}} — {{description}}, measuring programmatic SEO performance' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Reddit Posts Source: https://maestrix.ai/skills/create-reddit-posts > **EXECUTION RULE**: Generate the complete set of Reddit post drafts yourself, then render them as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Generate 5-8 Reddit-native post drafts tailored to specific subreddits. Each draft follows a format that performs on Reddit (story, data drop, question, AMA, discussion) and avoids anything that reads like marketing. Output includes the post body, predicted engagement, and the rationale for the chosen format per subreddit. ## CONTEXT NEEDED Before generating, the user should provide: - **Product / brand name**: e.g. "Maestrix" - **One-liner**: what it does - **Target subreddits**: 3-6 subreddits (if unknown, ask the user to run `analyze_reddit_opportunity` first) - **Goal per subreddit** *(optional)*: discussion / brand awareness / soft traffic / community building - **Founder story or unusual data** *(optional)*: fuels the "story" and "data drop" formats - **Account age + karma** *(optional)*: gates how directly the brand can be mentioned ## REDDIT POST FORMAT CATALOG The user should get a mix of these formats across the 5-8 drafts. Each format has its own structural rules. ### Format A — Story Post - **Hook line**: ≤15 words, first-person, conflict-led - **Setup**: 80-120 words establishing the situation - **Turn**: 60-100 words on the unexpected moment or realization - **Payoff**: 40-80 words with the lesson or open question to the community - **Brand mention**: only if context demands it; otherwise zero ### Format B — Data Drop - **Headline**: ≤12 words stating the surprising finding - **Methodology paragraph**: 60-80 words on how the data was collected - **3-5 findings**: each as a bullet with the number + 1-sentence interpretation - **Caveats**: 2-3 honest limitations - **Open question**: invite the community to challenge or extend the data ### Format C — Question Post - **Question line**: ≤15 words, specific enough to be answerable, broad enough to invite many voices - **Context paragraph**: 60-100 words explaining why you're asking and what you've already tried - **Constraints**: 2-3 bullets on what answers WOULDN'T help (filters noise) - **What you'll do with answers**: 1 sentence commitment back to the community ### Format D — AMA - **Title**: "I'm [role/credential]. AMA about [specific topic]." ≤80 chars - **Opening post**: 150-200 words establishing why you're qualified + 3 things you're happy to dig into - **Verification plan**: how you'll prove identity to the mods - **Seed questions**: 5 questions the brand should answer first to set tone - **Promotion rules**: what the brand will and will NOT mention ### Format E — Discussion / Opinion - **Hot take headline**: ≤15 words, mildly contrarian - **Position**: 100-140 words defending the take - **3 counterarguments**: steel-manned, each 30-40 words - **Response to each**: why the position still holds - **CTA**: ask the community what they'd add ## DELIVERABLE STRUCTURE For each of the 5-8 drafts, output: - **Target subreddit** + chosen format + rationale - **Title** (final, ready to paste) - **Body** (final, ready to paste, formatted with Reddit-native markdown) - **Best posting day + time** (specify timezone) - **Predicted engagement signal**: upvote range + comment range based on subreddit averages - **Mod risk score**: low / medium / high + why - **Repurpose path**: how to recycle this post as a tweet thread, LinkedIn post, or newsletter blurb ## QUALITY RULES 1. NO opening commentary — start with the first draft 2. NO marketing speak — no "solution", "leverage", "unlock", "game-changer", "revolutionize" 3. NO brand mention in the first 80% of any story or data post 4. Every title passes the "would a real human upvote this" test 5. Formatting uses Reddit-native markdown (no emoji headers, no walls of text, paragraph breaks every 2-3 sentences) 6. Mix at least 3 distinct formats across the 5-8 drafts 7. Each draft is paste-ready, not a template with [BRACKETS] --- ## UI Rendering Contract **Pattern:** F (Editorial Document) **Layout spec:** Document header (brand name + "Reddit Post Drafts" + total draft count + format mix chips). Stack of draft cards, one per post. Each card: subreddit badge top-left, format chip top-right, title at 18px, body in monospace-feel block with Reddit-style formatting preserved, metadata strip at bottom (best time, engagement prediction, mod risk). Copy-to-clipboard affordance per draft. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save these Reddit posts to my {{companyName}} project" - **find_reddit_threads** (green, primary) — sendPrompt: "Find active Reddit threads for {{companyName}} to comment on this week" - **analyze_reddit_opportunity** (neutral, secondary) — sendPrompt: "Map more Reddit opportunities for {{companyName}} — new subreddits we haven't tested" - **create_social_posts** (neutral, secondary) — sendPrompt: "Repurpose these Reddit posts for {{companyName}} into LinkedIn and X content" - **create_content_distribution** (neutral, secondary) — sendPrompt: "Build the distribution loop for {{companyName}} around Reddit + email + LinkedIn" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Create Referral Program Source: https://maestrix.ai/skills/create-referral-program > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Design a complete referral program with incentive structures, viral mechanics, and implementation strategy. ## CONTEXT NEEDED - Company name and description - Product/service type - Current customer LTV - Customer acquisition cost (CAC) - Target referral rate - Primary customer motivation ## REFERRAL PROGRAM FRAMEWORK ### 1. Program Foundation Analysis **Business Metrics** | Metric | Current | Target | Impact on Program | |--------|---------|--------|-------------------| | Customer LTV | $ | $ | | | Customer CAC | $ | $ | | | LTV:CAC Ratio | | 3:1+ | | | Avg. transaction value | $ | $ | | | Purchase frequency | | | | | Churn rate | % | % | | **Referral Budget Calculation** - Max incentive budget per referral: $[LTV × margin - target CAC] - Recommended referrer reward: $[X] - Recommended referee reward: $[Y] - Program ROI target: [X]% ### 2. Customer Motivation Analysis **Referrer Motivations** | Motivation Type | Applies? | Incentive Strategy | |-----------------|----------|-------------------| | Financial gain | | | | Status/recognition | | | | Helping friends | | | | Exclusive access | | | | Community building | | | **Referee Motivations** | Motivation Type | Applies? | Entry Incentive | |-----------------|----------|-----------------| | Discount/savings | | | | Extended trial | | | | Premium features | | | | Exclusive access | | | | Social proof | | | ### 3. Incentive Structure Design **Recommended Program Type** | Type | Pros | Cons | Fit Score | |------|------|------|-----------| | Double-sided | Both parties rewarded | Higher cost | | | Single-sided | Simple, lower cost | Less viral | | | Tiered | Scales with activity | Complex | | | Milestone-based | Gamification | Hard to track | | | Credit-based | Flexible redemption | Liability | | **Incentive Recommendation** ``` REFERRER GETS: [Specific reward] - Trigger: [When reward is given] - Value: $[Amount or equivalent] - Expiration: [If applicable] REFEREE GETS: [Specific reward] - Trigger: [When reward is given] - Value: $[Amount or equivalent] - Expiration: [If applicable] ``` **Tiered Structure (if applicable)** | Tier | Referrals Required | Reward | Status Unlocked | |------|-------------------|--------|-----------------| | Bronze | 1-2 | | | | Silver | 3-5 | | | | Gold | 6-10 | | | | Platinum | 11+ | | | ### 4. Viral Loop Design **K-Factor Optimization** ``` K = i × c Where: i = Invites per user = [Target] c = Conversion rate = [Target] K-factor target = [>1 for viral growth] ``` **Viral Loop Mapping** ``` [Trigger Moment] → [Share Action] → [Invite Received] → [Signup] → [Activation] → [New Trigger] ↓ ↓ ↓ ↓ ↓ [Timing] [Channels] [Messaging] [Reward] [Repeat] ``` **Sharing Triggers (Aha Moments)** | Trigger Event | Timing | Share Prompt | Expected Response | |---------------|--------|--------------|-------------------| | After first success | Immediate | | | | After milestone | Achievement | | | | During positive experience | In-moment | | | | After support resolution | Post-help | | | ### 5. Channel Strategy **Sharing Channels** | Channel | Ease | Reach | Conversion | Priority | |---------|------|-------|------------|----------| | Email | | | | | | SMS/Text | | | | | | Social (LinkedIn) | | | | | | Social (Twitter/X) | | | | | | Direct link | | | | | | In-app share | | | | | **Channel-Specific Messaging** | Channel | Message Template | Character Limit | CTA | |---------|-----------------|-----------------|-----| | Email | | | | | SMS | | 160 | | | LinkedIn | | | | | Twitter | | 280 | | ### 6. Program Mechanics **Tracking & Attribution** - Unique referral codes: [Format] - Link structure: [domain.com/refer/CODE] - Attribution window: [X] days - Multi-touch attribution: [Yes/No] **Rules & Limitations** | Rule | Setting | Rationale | |------|---------|-----------| | Max referrals per user | | | | Self-referral prevention | | | | Referral expiration | | | | Minimum transaction | | | | Geographic restrictions | | | | Fraud prevention | | | ### 7. User Experience Flow **Referrer Journey** ``` 1. [Discovery] - Where they learn about program 2. [Activation] - How they get their link/code 3. [Sharing] - How they invite friends 4. [Tracking] - How they see progress 5. [Reward] - How they receive incentive 6. [Repeat] - How they're encouraged to continue ``` **Referee Journey** ``` 1. [Receive] - How they get the invitation 2. [Click] - Landing page experience 3. [Signup] - Registration with referral tracking 4. [Qualify] - Actions to trigger rewards 5. [Reward] - Receiving their incentive 6. [Convert] - Becoming a referrer themselves ``` ### 8. Messaging Framework **Program Naming** | Option | Tone | Clarity Score | |--------|------|---------------| | [Name] Ambassador Program | Professional | | | [Name] Rewards | Benefit-focused | | | Give $X, Get $X | Direct value | | | Friends & Family | Personal | | **Key Messages** **Headline Options** 1. [Benefit-focused headline] 2. [Value-focused headline] 3. [Community-focused headline] **Value Proposition** ``` Give: [What referrer shares] Get: [What referrer receives] Friend Gets: [What referee receives] ``` **Email/Share Templates** | Scenario | Subject/Hook | Body | CTA | |----------|--------------|------|-----| | Initial invite | | | | | Reminder | | | | | Reward notification | | | | | Milestone achieved | | | | ### 9. Implementation Roadmap **Phase 1: MVP Launch (Week 1-2)** - [ ] Basic tracking system - [ ] Unique referral links - [ ] Landing page - [ ] Email notifications - [ ] Manual reward fulfillment **Phase 2: Automation (Week 3-4)** - [ ] Automated reward triggers - [ ] Referral dashboard - [ ] Analytics integration - [ ] Fraud detection basics **Phase 3: Optimization (Month 2)** - [ ] A/B testing framework - [ ] Advanced analytics - [ ] Tiered rewards - [ ] Gamification elements **Phase 4: Scale (Month 3+)** - [ ] API integrations - [ ] Partner programs - [ ] Influencer tiers - [ ] Cross-promotion ### 10. Success Metrics **Primary KPIs** | Metric | Target | Measurement | |--------|--------|-------------| | Referral rate | % of customers who refer | | | Conversion rate | % of invites that convert | | | K-factor | Viral coefficient | | | Referral CAC | Cost per referred customer | | | Referral LTV | Lifetime value of referred | | **Secondary Metrics** | Metric | Target | Measurement | |--------|--------|-------------| | Invites per referrer | | | | Time to first referral | | | | Share rate by channel | | | | Reward redemption rate | | | ### 11. Platform Recommendations | Platform | Best For | Integration Effort | Monthly Cost | |----------|----------|-------------------|--------------| | Rewardful | SaaS/subscription | Low | $29+ | | ReferralCandy | E-commerce | Low | $59+ | | Friendbuy | Enterprise | Medium | Custom | | GrowSurf | B2B SaaS | Low | $0-$200 | | Build custom | Full control | High | Dev cost | ## QUALITY RULES 1. NO opening comments - start with analysis 2. Calculate actual incentive economics with real numbers 3. Design for K-factor > 1 viral potential 4. Include specific messaging templates 5. Provide implementation timeline with phases ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a referral program landing page for {{companyName}} — {{description}}' Secondary actions (neutral styling): - Label: 'Lifecycle emails ↗' sendPrompt: 'Create email sequences for {{companyName}} — {{description}}, promoting the referral program' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, launching the referral program' - Label: 'Campaign ideas ↗' sendPrompt: 'Generate campaign ideas for {{companyName}} — {{description}}, to boost referral signups' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Sales Script Source: https://maestrix.ai/skills/create-sales-script > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create comprehensive sales scripts for discovery calls, demos, and closing conversations. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience/buyer persona - Sales cycle stage (discovery, demo, closing) - Common objections (optional) ## SALES SCRIPT FRAMEWORK ### 1. Call Preparation **Pre-Call Research Checklist** - [ ] Company background - [ ] Recent news/announcements - [ ] LinkedIn profile of contact - [ ] Current tech stack (if visible) - [ ] Previous interactions - [ ] Potential pain points - [ ] Competition they might use ### 2. Discovery Call Script (45-60 min) **Opening (2-3 min)** "Hi [Name], thanks for taking the time today. Before we dive in, I'd love to learn more about [Company] and what prompted you to take this call. Then I can share how we might be able to help. Sound good?" **Rapport Building (3-5 min)** - Acknowledge their time - Reference research findings - Establish mutual connection if applicable **Situation Questions (10 min)** 1. "Can you walk me through how your team currently handles [process]?" 2. "What tools or systems are you using today for [area]?" 3. "How many people are involved in this process?" 4. "What does success look like for your team?" **Problem Questions (10 min)** 1. "What's the biggest challenge you're facing with [area]?" 2. "How long has this been an issue?" 3. "What have you tried so far to solve it?" 4. "What happens if nothing changes?" **Impact Questions (10 min)** 1. "How is this affecting [metric/goal]?" 2. "What's the cost of this problem in time/money?" 3. "Who else is impacted by this issue?" 4. "What would solving this mean for you personally?" **Vision Questions (5 min)** 1. "If you could wave a magic wand, what would the ideal solution look like?" 2. "What would need to be true for you to move forward?" **Next Steps (5 min)** - Summarize what you heard - Confirm fit - Propose next step ### 3. Demo Script (30-45 min) **Opening (5 min)** - Recap discovery findings - Confirm agenda - Set expectations **Tailored Demo Flow** | Their Pain Point | Feature to Show | Business Outcome | |------------------|-----------------|------------------| | [Pain 1] | [Feature] | [Outcome] | | [Pain 2] | [Feature] | [Outcome] | | [Pain 3] | [Feature] | [Outcome] | **Demo Narration Template** "You mentioned [pain point]. Let me show you how [feature] works. [Demo action]. What this means for you is [benefit]. Companies like [reference] have seen [result]." **Engagement Checkpoints** Every 5-7 minutes: - "How does that compare to your current process?" - "Can you see your team using this?" - "Any questions so far?" **Closing the Demo** - Summary of value shown - ROI projection - Next steps ### 4. Objection Handling Library **Price Objections** "It's too expensive." → "I understand budget is a consideration. Let me ask—if we could demonstrate ROI of [X], would the investment make sense? Let's look at the numbers together..." **Timing Objections** "Not the right time." → "I hear you. Can you help me understand what would need to change for timing to be right? Often our customers felt the same way but found that [reason to act now]..." **Competition Objections** "We're looking at [Competitor]." → "Great, they're a solid company. What specifically are you comparing? Where we typically see the biggest difference is [differentiator]..." **Authority Objections** "I need to talk to my team." → "Absolutely. Would it help if I provided materials you could share? What questions do you think they'll have?" **Status Quo Objections** "We're fine with what we have." → "That makes sense—you've made it work. Curious though, you took this meeting for a reason. What were you hoping to learn?" ### 5. Closing Script **Trial Close** "Based on everything we've discussed, do you see this solving [main problem]?" **Proposal Walkthrough** - Investment summary - Implementation timeline - Success metrics - Support included **Assumptive Close** "Let's talk about getting started. What would the implementation timeline look like on your end?" **Urgency Elements** - Limited capacity - Pricing validity - Upcoming changes ### 6. Voicemail Scripts (30 seconds) **Initial Outreach** "Hi [Name], this is [Your name] from [Company]. I noticed [personalized observation] and thought you'd find [specific value] interesting. My number is [number]. Again, that's [Your name] from [Company] at [number]." **Follow-Up** "Hi [Name], [Your name] again from [Company]. Following up on [previous touchpoint]. Quick question: [intriguing question]? Call me at [number] or reply to my email." ### 7. Email Follow-Up Templates **Post-Discovery** **Post-Demo** **Post-Proposal** **No Response** ## QUALITY RULES 1. NO opening comments 2. Natural conversation, not reading scripts 3. Questions before pitching 4. Always tie to their specific situation ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Pipeline playbook ↗' - sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, incorporating this sales script into the process' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, using the talk track from this script' - Label: 'Sales roleplay ↗' sendPrompt: 'Run a sales roleplay for {{companyName}} — {{description}}, practicing the objection handling from this script' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a leave-behind resource for this sales conversation' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Seo Strategy Source: https://maestrix.ai/skills/create-seo-strategy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Perform a comprehensive keyword gap analysis to enhance online visibility and competitiveness. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Current keywords (optional) ## SEO STRATEGY FRAMEWORK ### 1. Keyword Research & Data Collection Identify **30 high-impact keywords** across six types: | Keyword | Type | Search Volume | Difficulty | Competition | Competitors | Strategic Asset | Target Persona | Title | Distribution | Success Metrics | |---------|------|---------------|------------|-------------|-------------|-----------------|----------------|-------|--------------|-----------------| **Keyword Types:** - Transactional (purchase intent) - Informational (learning intent) - Commercial (comparison intent) - LSI (related semantic terms) - Long-tail (specific phrases) - People Also Asked (question queries) ### 2. Competitor Analysis - Top-ranking pages for each keyword - Their content strategies - Backlink profiles - Schema usage ### 3. Strategic Asset Recommendations For each keyword, recommend: - Content type (guide, calculator, tool, video) - Multimedia formats - Distribution channels ### 4. SEO-Optimized Titles - Actionable hooks and CTAs for each keyword ### 5. Target Personas - Match keywords to audience segments ### 6. Success KPIs - Organic traffic, CTR, conversion rate targets ## PROGRAMMATIC SEO OPPORTUNITIES Identify 5+ template opportunities for scalable SEO pages: | Template Type | URL Pattern | Example | Pages Potential | Traffic Estimate | |---------------|-------------|---------|-----------------|------------------| | Location | /[service]-[city] | /marketing-agency-austin | 500+ | 50k/mo | | Integration | /integrations/[tool] | /integrations/salesforce | 50+ | 10k/mo | | Comparison | /[brand]-vs-[competitor] | /hubspot-vs-salesforce | 20+ | 15k/mo | | Use-case | /[use-case]-software | /email-marketing-software | 30+ | 25k/mo | For each template, provide: - Title tag template with keyword variables - H1 template - Key content sections required - Internal linking strategy - Unique value per page ## TECHNICAL SEO AUDIT | Element | Check | Priority | |---------|-------|----------| | Canonical tags | Correct implementation | High | | Core Web Vitals | LCP, FID, CLS scores | High | | Mobile-first indexing | Mobile rendering | High | | XML sitemap | Complete and submitted | Medium | | Structured data | Schema markup present | Medium | | Internal linking depth | Max 3 clicks to content | Medium | ## CONTENT DECAY ANALYSIS Identify refresh opportunities: | Decay Signal | Action | Priority | |--------------|--------|----------| | Dated statistics (>2 years) | Update with current data | High | | Outdated screenshots | Replace with current UI | Medium | | Old pricing/features | Verify accuracy | High | | Declining traffic | Refresh + republish | High | ## QUALITY RULES 1. NO opening comments - start directly with content 2. Use REAL, verifiable search volume data with sources 3. NO hallucinated statistics 4. Provide actionable recommendations 5. Include pSEO template opportunities ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B **Layout spec:** Persistent header (metric scores/bars) → tab bar (3-5 tabs, --primary underline active) → per-tab sub-layouts: summary (metric cards), detail (card list), distribution (bar chart rows), sources (tag clusters) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Programmatic SEO ↗' - sendPrompt: 'Create a programmatic SEO plan for {{companyName}} — {{description}}, based on this SEO strategy' Secondary actions (neutral styling): - Label: 'LLM optimization ↗' sendPrompt: 'Create an LLM optimization plan for {{companyName}} — {{description}}, complementing this SEO strategy' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, aligned with the keyword strategy we just built' - Label: 'Article from topic ↗' sendPrompt: 'Write an article for {{companyName}} — {{description}}, targeting the top priority keyword from this strategy' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Shopping Ads Source: https://maestrix.ai/skills/create-shopping-ads > **EXECUTION RULE**: Generate the complete output yourself, then render it as an interactive HTML widget using `show_widget`. Do not return raw markdown. Use only hardcoded hex colors from the Maestrix design system. Never use CSS variables. Never make a secondary API call. Build production-ready Google Shopping + Performance Max content: optimized product titles, feed descriptions, custom labels, asset group headlines/descriptions, audience signals, and bidding recommendations. ## CONTEXT NEEDED - Company name and description - Product name + category - Product features - Price point - Target audience / markets - Competitor pricing (optional) ## SHOPPING ADS FRAMEWORK ### 1. Product Feed Optimization - 5 optimized product title variations (70 chars each): primary, feature-focused, benefit-focused, audience-specific, seasonal/promotional - Feed description (5000 chars max) with front-loaded keywords, materials, sizes, use cases, brand story - Google product category mapping (recommended path + alternatives) ### 2. Custom Labels Strategy - Label 0: Profit margin tiers (high/medium/low) - Label 1: Performance tiers (best sellers / good performers / new / clearance) - Label 2: Seasonality (evergreen / seasonal_q4 / trending) - Label 3: Price competitiveness (price_leader / competitive / premium) - Label 4: Stock status (in_stock / limited / preorder) ### 3. Performance Max Asset Groups **Asset Group 1 — High-Intent Converters** - Audience signals (custom segments, your data: cart abandoners, past purchasers) - 15 short headlines (30 chars): 5 feature, 5 benefit, 5 urgency - 5 long headlines (90 chars) - 5 descriptions (90 chars) **Asset Group 2 — Awareness / Consideration** - Audience signals (in-market + affinity) - Headlines & descriptions (awareness messaging) ### 4. Bidding & Budget Recommendations - Recommended daily budget by category - Target ROAS by asset group - Campaign priority settings (low/medium/high) - Negative keywords list (brand misspellings, irrelevant terms, optional competitor brands) ### 5. Feed Quality Checklist - GTIN/MPN included - High-resolution images (min 800x800) - Accurate shipping & tax - Product ratings enabled - Promotion feed configured ## QUALITY RULES 1. NO opening fluff 2. All character limits respected (30/70/90/200) 3. Real keyword density in titles 4. Audience signals are specific, not generic 5. Each headline tests a distinct angle (no near-duplicates) ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract **Pattern:** D (Sequential Steps) Numbered step cards: Feed Optimization → Custom Labels → Asset Groups → Bidding → Quality Check. Dashed connectors. Each card includes type badge + chips. Bottom summary with total assets generated, recommended ROAS target, launch checklist. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ### What's next actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this shopping ads plan to my {{companyName}} project" - **create_meta_ads** (green, primary) — sendPrompt: "Create Meta ads for {{companyName}} to complement the Google Shopping campaign" - **google_merchant_feed** (neutral, secondary) — sendPrompt: "Build a Google Merchant Center feed for {{companyName}}" - **pdp_audit** (neutral, secondary) — sendPrompt: "Audit the product detail page for {{companyName}} to lift Quality Score" - **create_landing_page** (neutral, secondary) — sendPrompt: "Create a landing page for {{companyName}} optimized for shopping ad traffic" --- ## Skill: Create Signup Flow Cro Source: https://maestrix.ai/skills/create-signup-flow-cro > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Optimize signup and registration flows for maximum conversion using behavioral psychology and UX best practices. ## CONTEXT NEEDED - Current signup flow (steps, fields) - Signup method (email, social, SSO) - Product type (SaaS, e-commerce, marketplace) - Current conversion rate (if known) - Post-signup onboarding goals - Target audience ## SIGNUP FLOW OPTIMIZATION FRAMEWORK ### 1. Current Flow Audit **Flow Mapping** ``` [Entry Point] → [Step 1] → [Step 2] → [Step 3] → [Confirmation] ↓ ↓ ↓ ↓ ↓ [Drop-off] [Drop-off] [Drop-off] [Drop-off] [Success] X% X% X% X% X% ``` **Field-by-Field Analysis** | Field | Required? | Type | Purpose | Friction Level | Remove/Keep/Defer? | |-------|-----------|------|---------|----------------|-------------------| | Email | | | | | | | Password | | | | | | | Name | | | | | | | Company | | | | | | | Phone | | | | | | | [Other] | | | | | | **Friction Score Assessment** | Friction Factor | Current State | Score (1-5) | Fix Priority | |-----------------|---------------|-------------|--------------| | Number of fields | | | | | Number of steps | | | | | Password requirements | | | | | Verification needed | | | | | Error handling | | | | | Mobile optimization | | | | | Load time | | | | ### 2. Form Field Optimization **Minimum Viable Signup (MVS)** ``` Essential fields only for first signup: ├── Email (account identifier) ├── Password OR │ └── Social auth (Google, Microsoft, etc.) └── [1 optional context field if critical] Defer to post-signup: ├── Full name ├── Company name ├── Role/title ├── Phone number ├── Company size ├── Use case └── How did you hear about us ``` **Progressive Profiling Strategy** | Timing | Fields to Collect | Why Now | UX Pattern | |--------|-------------------|---------|------------| | Signup | Email, password | Account creation | Required | | First login | Name | Personalization | Welcome modal | | After value | Company, role | Segmentation | Sidebar prompt | | Before upgrade | Phone, size | Sales readiness | Inline form | **Field Design Best Practices** | Best Practice | Implementation | |---------------|----------------| | Inline validation | Show errors as user types | | Clear error states | Red border + specific message | | Smart defaults | Pre-select common options | | Auto-formatting | Phone, credit card masks | | Placeholder + label | Don't use placeholder as label | | Logical tab order | Match visual flow | | Autofill support | Proper input types | ### 3. Password Optimization **Password UX Improvements** | Pattern | Description | Impact | |---------|-------------|--------| | Show/hide toggle | Eye icon to reveal password | High | | Strength meter | Visual password strength | Medium | | Real-time validation | Show requirements met | High | | Copy-paste enabled | Don't block paste | Medium | | Reasonable requirements | 8+ chars, avoid complex rules | High | **Alternative Auth Strategies** | Method | Conversion Impact | Best For | |--------|-------------------|----------| | Social login (Google) | +15-30% | B2B SaaS | | Magic link (passwordless) | +10-20% | Consumer apps | | SSO/SAML | Required | Enterprise | | Phone OTP | +5-10% | Mobile-first | ### 4. Multi-Step Flow Optimization **Single-Step vs. Multi-Step Decision** | Factor | Single-Step Better | Multi-Step Better | |--------|-------------------|-------------------| | Fields needed | <4 fields | >5 fields | | Complexity | Simple product | Complex onboarding | | Intent clarity | High intent visitors | Exploratory visitors | | Mobile usage | Low mobile | High mobile | **Multi-Step Best Practices** ``` Step 1: Account (Email + Password) ↓ [Progress: ●○○] Step 2: Profile (Name + Company) ↓ [Progress: ●●○] Step 3: Personalization (Role + Use case) ↓ [Progress: ●●●] [Dashboard with contextual onboarding] ``` **Progress Indicators** | Pattern | When to Use | |---------|-------------| | Step dots | 2-4 steps | | Progress bar | 4+ steps or percentage complete | | Step names | When steps have clear themes | | None | 2 steps with fast completion | ### 5. Trust & Anxiety Reduction **Trust Signals for Signup** | Element | Placement | Example | |---------|-----------|---------| | Security badge | Near password field | 256-bit encryption | | Privacy assurance | Below email field | "We'll never spam you" | | Customer logos | Above/beside form | 5+ recognizable logos | | User count | Near CTA | "Join 50,000+ teams" | | No credit card | CTA area | "No credit card required" | **Objection Handling** | Objection | Pre-emption | |-----------|-------------| | "Too much info" | "Takes less than 60 seconds" | | "Will I be spammed?" | "We only send what you need" | | "Is it secure?" | Security badges + privacy link | | "Can I cancel?" | "Cancel anytime, no questions" | ### 6. CTA Optimization **CTA Copy Framework** | Pattern | Example | Best For | |---------|---------|----------| | Action + Benefit | "Start building for free" | SaaS products | | Simple action | "Create account" | Broad audiences | | Value focused | "Get started free" | Low-commitment | | Specific outcome | "Get your report" | Lead magnets | **CTA Design Principles** - High contrast with background - Full-width on mobile - Clear hover/active states - Loading state on submit - Disabled state until valid ### 7. Social Proof Integration **Social Proof Placement** | Location | Type | Impact | |----------|------|--------| | Above form | Customer logos | Credibility | | Beside form | Testimonial quote | Trust | | Below CTA | User count | Bandwagon | | Inline | Star rating | Quality signal | **Dynamic Social Proof** ```javascript // Real-time activity indicators "Sarah from Austin just signed up" "243 people signed up this week" "3 people viewing this page" ``` ### 8. Error Prevention & Handling **Inline Validation Rules** | Field | Validate On | Error Message | Success Message | |-------|-------------|---------------|-----------------| | Email | Blur | "Please enter a valid email" | ✓ | | Password | Keyup | "Password must be 8+ characters" | Strength meter | | Confirm password | Keyup | "Passwords don't match" | ✓ Match | **Error State Design** - Red border (not just text color) - Icon indicator (⚠️) - Specific, helpful message - Positioned near the field - Clear path to fix ### 9. Mobile Optimization **Mobile-Specific Improvements** | Issue | Solution | |-------|----------| | Small tap targets | Min 48px height for inputs | | Keyboard overlap | Scroll form into view | | Wrong keyboard | Use correct input types | | Slow typing | Enable autocomplete/autofill | | Fat finger errors | Adequate spacing between elements | **Input Types for Mobile** | Field | Input Type | Mobile Benefit | |-------|------------|----------------| | Email | `type="email"` | @ key visible | | Phone | `type="tel"` | Number pad | | URL | `type="url"` | .com key | | Number | `type="number"` | Number pad | ### 10. Post-Signup Optimization **Confirmation Page Strategy** | Element | Purpose | Example | |---------|---------|---------| | Success message | Confirmation | "You're in! Check your email" | | Next step CTA | Activation | "Complete your profile" | | Value reminder | Motivation | "Here's what you can do now" | | Help access | Support | "Questions? Chat with us" | **Email Verification Flow** | Approach | Pros | Cons | Best For | |----------|------|------|----------| | Verify before access | Clean list | Friction | Strict compliance | | Verify within 24h | Low friction | Some bounces | Most SaaS | | No verification | Zero friction | List quality | Consumer apps | ### 11. A/B Test Roadmap **High-Impact Tests** | Test | Hypothesis | Expected Impact | Priority | |------|------------|-----------------|----------| | Add Google SSO | Reduces friction | +15-25% | Critical | | Remove company field | Fewer fields = higher completion | +10-15% | High | | Single-step vs multi | [depends on context] | +/-15% | High | | Social proof addition | Trust increases signup | +5-10% | Medium | | CTA copy variations | Better copy = more clicks | +5-10% | Medium | ### 12. Metrics & Measurement **Funnel Metrics** | Stage | Metric | Target | Current | |-------|--------|--------|---------| | Page views | Visitors to signup page | | | | Started | Began form interaction | | | | Completed | Submitted form | | | | Verified | Email confirmed | | | | Activated | Completed key action | | | **Conversion Rate Benchmarks** | Product Type | Signup Rate | Activation Rate | |--------------|-------------|-----------------| | B2B SaaS | 2-5% | 20-40% | | Consumer app | 5-15% | 10-30% | | E-commerce | 3-8% | varies | | Freemium | 8-15% | 15-25% | ## QUALITY RULES 1. NO opening comments - start with flow audit 2. Minimize required fields (email + password at most) 3. Include specific UX recommendations 4. Consider mobile-first design 5. Provide measurable improvement targets ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** H **Layout spec:** Score header (large number, color-coded: >80 --success, 60-80 --warning, <60 --destructive, sub-score chips) → ranked recommendation list (severity badge, title, effort/impact chips) → quick wins (--success tint) → implementation timeline (Now/30d/90d). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'A/B testing plan ↗' - sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to optimize the signup flow' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create an optimized landing page for {{companyName}} — {{description}}, based on signup flow insights' - Label: 'Lifecycle emails ↗' sendPrompt: 'Create onboarding emails for {{companyName}} — {{description}}, to improve post-signup activation' - Label: 'Analytics tracking ↗' sendPrompt: 'Create an analytics tracking plan for {{companyName}} — {{description}}, measuring signup funnel metrics' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Social Posts Source: https://maestrix.ai/skills/create-social-posts > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create engaging social media content optimized for each platform's best practices. ## CONTEXT NEEDED - Company name and description - Content topic or campaign - Target audience - Platforms (LinkedIn, Twitter/X, Instagram, Facebook, TikTok) - Tone preference (optional) ## SOCIAL POST FRAMEWORK ### 1. Content Pillars Alignment Map this content to: - Pillar category (Educational, Promotional, Entertainment, Community) - Content goal (Awareness, Engagement, Traffic, Conversion) - Buyer journey stage ### 2. LinkedIn Posts (5 Variations) **Post 1: Story-Based** [Hook line - pattern interrupt] [Personal story or observation - 2-3 paragraphs] [Key insight or lesson] [Call to action] [Hashtags: 3-5 relevant] **Post 2: List/Carousel Format** [Hook with number] [Slide 1: Title] [Slide 2-9: One point each] [Slide 10: CTA] Caption: [Supporting text] **Post 3: Contrarian Take** **Post 4: Question-Led** **Post 5: Data/Stat-Led** ### 3. Twitter/X Posts (10 Variations) **Thread Format** (if applicable) 1/ [Hook tweet - controversial or intriguing] 2/ [Context] 3-8/ [Main points] 9/ [Summary] 10/ [CTA with link] **Single Tweets** - Tweet 1: [Question format] - Tweet 2: [Hot take] - Tweet 3: [Stat + insight] - Tweet 4: [Before/after] - Tweet 5: [Analogy] - Tweet 6: [Listicle] - Tweet 7: [Quote-style] - Tweet 8: [Meme format text] ### 4. Instagram Posts (3 Variations) **Post 1: Carousel** - Slide concepts (10 slides) - Caption (2200 chars max with CTA) - Hashtags (20-30, in comment) **Post 2: Single Image** - Image concept - Caption - Hashtags **Post 3: Reel Script** - Hook (0-3 sec) - Setup (3-10 sec) - Content (10-50 sec) - CTA (last 5 sec) - Caption - Hashtags - Audio suggestion ### 5. Facebook Posts (3 Variations) **Post 1: Long-form** **Post 2: Video intro script** **Post 3: Community question** ### 6. TikTok Scripts (3 Variations) **Script 1** - Hook (0-2 sec): [Attention grabber] - Content (3-45 sec): [Main value] - CTA (end): [Follow/comment prompt] - On-screen text - Audio/sound suggestion - Trending format reference **Script 2** **Script 3** ### 7. Cross-Platform Adaptations | Core Message | LinkedIn | Twitter | Instagram | TikTok | |--------------|----------|---------|-----------|--------| | [Message 1] | [Adapted] | [Adapted] | [Adapted] | [Adapted] | | [Message 2] | | | | | | [Message 3] | | | | | ### 8. Hashtag Strategy **Brand Hashtags**: [3] **Industry Hashtags**: [10] **Trending/Niche Hashtags**: [10] Platform-specific recommendations. ### 9. Engagement Tactics - Reply prompts to include - Comment response templates - Community interaction plan ### 10. Posting Recommendations | Platform | Best Day | Best Time | Frequency | |----------|----------|-----------|-----------| ## QUALITY RULES 1. NO opening comments 2. Platform-specific character limits respected 3. Native format for each platform 4. Engagement-focused (not just broadcast) ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, expanding on the social post themes' Secondary actions (neutral styling): - Label: 'Blog article ↗' sendPrompt: 'Write a blog article for {{companyName}} — {{description}}, expanding on the best-performing social post topic' - Label: 'Email sequence ↗' sendPrompt: 'Create a lifecycle email sequence for {{companyName}} — {{description}}, aligned with these social themes' - Label: 'Campaign ideas ↗' sendPrompt: 'Generate campaign ideas for {{companyName}} — {{description}}, building on these social post concepts' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Value Proposition Source: https://maestrix.ai/skills/create-value-proposition > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive value proposition ladder framework connecting features to capabilities to benefits to outcomes. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience segments ## VALUE LADDER FRAMEWORK ### Feature → Capability → Benefit → Outcome Mapping | Feature | Capability | Benefit | Outcome | |---------|------------|---------|---------| | [What you have] | [What it does] | [What customers get] | [What customers achieve] | ### Complete Value Ladder Analysis (5-7 ladders) For each: - **Feature**: Core functionality - **Capability**: What it enables - **Benefit**: Immediate value - **Outcome**: Long-term business impact - **Proof Point**: Evidence/metric ### Role-Based Value Propositions **Executive Level**: Strategic outcomes, ROI, risk mitigation **Manager Level**: Operational efficiency, team productivity **User Level**: Ease of use, time savings, job satisfaction ### Audience-Specific Messaging Matrix For each segment (3): - Context - Primary pain - Value ladder priority - Tailored value proposition ### Website Messaging Guide - Homepage hierarchy - Product page flow - Case study structure ### Proof Point Strategy - Quantitative proof (ROI, performance, time, cost) - Qualitative proof (testimonials, case studies, endorsements) ## QUALITY RULES 1. NO opening comments 2. Connect features to business outcomes 3. Use customer language, not jargon 4. Support all claims with proof points ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** E **Layout spec:** Top-level statement card (--primary/0.08 bg, 18px headline) → audience branch cards (label badge, core message, proof, example in --muted) → problem-solution pairs (2-col: --warning left, --success right) → VoC quotes → tone/language tag cloud. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, building on the value proposition we just defined' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, centered on this value proposition' - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, leading with this value proposition' - Label: 'Feature messaging ↗' sendPrompt: 'Create feature messaging for {{companyName}} — {{description}}, expanding on each value pillar' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create Video Script Source: https://maestrix.ai/skills/create-video-script > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create compelling video scripts optimized for different platforms and purposes. ## CONTEXT NEEDED - Company name and description - Video purpose (explainer, testimonial, ad, social, etc.) - Target audience - Key message - Video length target ## VIDEO SCRIPT FRAMEWORK ### 1. Video Strategy **Video Type**: [Explainer/Testimonial/Ad/Tutorial/Brand] **Platform**: [YouTube/LinkedIn/Instagram/TikTok/Website] **Length**: [Target duration] **Goal**: [Awareness/Consideration/Conversion] **CTA**: [Desired action] ### 2. Script Structure Options **Problem-Solution (60-90 sec)** - Hook (5 sec) - Problem (15 sec) - Agitation (10 sec) - Solution intro (15 sec) - How it works (20 sec) - Proof (10 sec) - CTA (5 sec) **Story Arc (2-3 min)** - Hook (5 sec) - Setup/Context (20 sec) - Challenge (30 sec) - Journey (45 sec) - Resolution (30 sec) - Key learning (15 sec) - CTA (10 sec) **Tutorial (3-5 min)** - Hook/Promise (10 sec) - Agenda (15 sec) - Step 1 (45 sec) - Step 2 (45 sec) - Step 3 (45 sec) - Recap (15 sec) - CTA (10 sec) ### 3. Complete Script **HOOK (0:00-0:05)** [Visual]: [Description] [Audio/VO]: "[Opening line - pattern interrupt or question]" [Text on screen]: [If applicable] **SCENE 1 (0:05-0:XX)** [Visual]: [Description] [Audio/VO]: "[Dialogue/narration]" [Music note]: [Mood/style] [B-roll suggestion]: [If applicable] [Continue for each scene...] **CTA (Final)** [Visual]: [Description] [Audio/VO]: "[Call to action]" [Text on screen]: [URL/action] ### 4. Alternative Scripts **Version A: Emotional Angle** [Full script] **Version B: Logical Angle** [Full script] **Version C: Social Proof Angle** [Full script] ### 5. Platform-Specific Adaptations **YouTube (Long-form)** - Hook within 5 seconds - Chapter markers - End screen strategy **LinkedIn (Professional)** - Text overlay for muted viewing - Professional tone - Business outcome focus **Instagram Reels/TikTok (Short-form)** - Vertical format notes - Trending audio consideration - Fast pacing - Text hooks **Website Hero** - Loop considerations - Autoplay without sound - Background vs. foreground ### 6. Visual Direction **Shot List** | Scene | Shot Type | Subject | Notes | |-------|-----------|---------|-------| | 1 | Wide | | | | 2 | Close-up | | | | 3 | B-roll | | | **Style Guidelines** - Color palette - Motion graphics style - Text treatment - Transition preferences ### 7. Audio Direction **Voiceover Notes** - Tone - Pacing - Emphasis points - Pronunciation guides **Music Direction** - Genre - Energy level - Mood - Reference tracks **Sound Effects** - Transition sounds - UI sounds (if showing product) - Ambient sounds ### 8. Testimonial Interview Questions **Opening** 1. "Tell us about yourself and your role." 2. "What was the situation before [Product]?" **Problem** 3. "What challenges were you facing?" 4. "What was the impact on your business?" **Solution** 5. "What made you choose [Product]?" 6. "How was the implementation process?" **Results** 7. "What results have you seen?" 8. "What specific metrics improved?" **Recommendation** 9. "What would you tell someone considering [Product]?" 10. "What's been the biggest surprise?" ### 9. A/B Testing Elements | Element | Version A | Version B | |---------|-----------|-----------| | Hook | | | | Thumbnail | | | | CTA | | | | Length | | | ### 10. Production Checklist - [ ] Script approved - [ ] Shot list complete - [ ] Talent confirmed - [ ] Location secured - [ ] Equipment ready - [ ] Graphics/animations brief - [ ] Music licensed - [ ] Captions prepared ## QUALITY RULES 1. NO opening comments 2. Visual and audio clearly separated 3. Timing notations included 4. Platform-optimized versions ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this video content' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, featuring this video' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, linking to this video' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building a series around this video' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Create White Paper Source: https://maestrix.ai/skills/create-white-paper > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create a comprehensive white paper with research-backed insights and actionable frameworks. ## CONTEXT NEEDED - Company name and description - White paper topic - Target audience - Key thesis or argument - Data sources available (optional) ## WHITE PAPER FRAMEWORK ### 1. White Paper Blueprint **Title Options** (5) - Research-style titles - Benefit-driven titles - Question-based titles **Subtitle** Supporting detail that adds specificity **Executive Summary** (250 words) - Problem context - Key findings - Main recommendations - Call to action ### 2. Document Structure **Cover Page** - Title, subtitle, author/company, date **Table of Contents** **Section Outline** 1. **Introduction** (500-750 words) - Hook/opening statement - Problem definition - Scope of research - Why this matters now 2. **Market Context** (750-1000 words) - Industry background - Current state analysis - Key trends - Challenges facing the industry 3. **Research Methodology** (300-500 words) - Data sources - Analysis approach - Limitations 4. **Key Findings** (2000-3000 words) - Finding 1: [Title] - Data/evidence - Analysis - Implications - Finding 2: [Title] - Finding 3: [Title] - Finding 4: [Title] - Finding 5: [Title] 5. **Framework/Model** (750-1000 words) - Proprietary framework introduction - Framework components - How to apply it - Visual representation 6. **Case Studies/Examples** (750-1000 words) - Example 1: Success story - Example 2: Lessons learned - Example 3: Industry application 7. **Recommendations** (750-1000 words) - Strategic recommendations - Tactical action items - Implementation roadmap 8. **Conclusion** (300-500 words) - Summary of insights - Future outlook - Call to action 9. **Appendix** - Additional data - Methodology details - Glossary ### 3. Research Requirements **Primary Research Needs** - Survey questions to ask - Interview topics - Data to collect **Secondary Research Sources** - Industry reports to cite - Academic sources - Government data - Analyst research ### 4. Visual Elements | Section | Visual Type | Purpose | Data Needed | |---------|-------------|---------|-------------| | Market Context | Chart | Show trend | Time series data | | Findings | Infographic | Summarize | Key stats | | Framework | Diagram | Explain model | Process flow | | Recommendations | Table | Compare options | Feature matrix | ### 5. Pull Quotes & Callouts Key statistics and quotes to highlight: - Stat callout 1 - Stat callout 2 - Expert quote 1 - Key insight box ### 6. Gated Content Strategy **Landing Page Copy** - Headline - Value proposition bullets - Form fields required - Trust signals **Email Nurture Sequence** - Email 1: Delivery + quick wins - Email 2: Deep dive on finding 1 - Email 3: Framework application - Email 4: Consultation offer ### 7. Promotion Plan **Content Derivatives** - Blog post series (5 posts) - Infographic - SlideShare deck - Webinar outline - Social snippets (20) - Email excerpts ## QUALITY RULES 1. NO opening comments 2. Research-backed claims with sources 3. Actionable frameworks, not just observations 4. Clear visual direction for designers ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Content plan ↗' - sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, using this white paper as a pillar piece' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, to gate this white paper download' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this white paper' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, offering this white paper as a lead magnet' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Creative Copy Source: https://maestrix.ai/skills/creative-copy > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Creative & Copy Professional-grade copywriting tools using proven frameworks (PAS, AIDA, How-To, Number, Question, Command) for high-converting marketing assets. ## CONTEXT NEEDED - **Company/product**: What you sell and who buys it - **Campaign theme or goal**: Launch, promotion, awareness, conversion - **Target audience**: Who will read this copy — role, pain, aspiration - **Tone**: Bold / witty / authoritative / empathetic / minimal - **Constraints** *(optional)*: Character limits, platform, keywords to include/avoid --- ## Available Tools (5) ### 1. generate_headlines Create high-converting headlines using proven formulas (PAS, AIDA, How-To, Number, Question, Command). ### 2. create_taglines Generate memorable brand taglines and slogans with strategic positioning. ### 3. generate_offers Create compelling promotional offers and value propositions. ### 4. create_marketing_copy Generate conversion-focused marketing copy for ads and campaigns. ### 5. generate_product_descriptions Create compelling product descriptions optimized for conversion and SEO. --- ## Output Rules - ✓ No opening comments or intros - ✓ No outros or meta-commentary - ✓ Direct content variations only - ✓ Framework identification (PAS, AIDA, etc.) labeled per output - ✓ A/B testing recommendations included --- ## Example Prompts 1. **Headlines**: "Generate 20 headlines for our productivity app launch" 2. **Taglines**: "Create tagline options for our eco-friendly cleaning brand" 3. **Offers**: "Design a compelling Black Friday offer for our SaaS" 4. **Marketing Copy**: "Write Facebook ad copy for our fitness app" 5. **Product Descriptions**: "Create product descriptions for our e-commerce store" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, adapted from this creative copy' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, using this creative direction' - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, with this creative angle' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building on this creative approach' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Customer Interview Source: https://maestrix.ai/skills/customer-interview > **EXECUTION RULE**: This is an INTERACTIVE SIMULATOR skill. Generate the interviewee persona card and play the conversation turn-by-turn in chat as a realistic human, then render the debrief as an interactive HTML widget using the `show_widget` tool. The widget IS the deliverable for the debrief phase. Do not break character mid-interview. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Run a realistic customer interview without scheduling, recording, or transcribing. The AI plays the interviewee with all the friction of a real conversation — vague answers, polite deflections, tangents — and then debriefs with the JTBD, pain hierarchy, exact phrases worth stealing for copy, and coaching on the moments the researcher missed. ## CONTEXT NEEDED Before the interview starts, the user should provide: - **Product / feature**: what you're researching — whole product, specific feature, pricing change, onboarding - **Interviewee persona**: current customer / churned user / prospect / lost deal — title, company size, tenure with product (or alternative they use) - **Research goal**: jobs-to-be-done / churn reason / feature validation / pricing sensitivity / onboarding friction / competitive switch - **What you already suspect** *(optional)*: hypothesis to pressure-test - **Interview style** *(optional)*: open-ended discovery / structured 5-question / "Mom Test" style / switch interview ## THE 3-PHASE FRAMEWORK ### Phase 1 — Interviewee Setup (rendered as a persona card before the interview) - **Identity**: name, title, company, tenure with product or alternative - **Current mood**: busy / cautious / venting / polite / skeptical (pick one, justify with 1 sentence) - **Hidden frustrations**: 3 frustrations they will NOT volunteer unless asked the right way - **What they think the interview is about**: usually different from the researcher's real goal - **What they will absolutely not admit on the first ask**: 1-2 things requiring trust to surface - **Vocabulary they use**: 8 in-character words/phrases (the AI must use these verbatim during the interview) ### Phase 2 — Live Interview (turn-by-turn in chat, AI stays in character) The AI plays the interviewee using these realistic behaviors: - Vague first answers — "it's fine, I guess" / "yeah we use it" - Tangents about adjacent tools or org politics - Polite deflections when the question is too direct ("good question, let me think...") - Sudden specificity when a question hits a real pain - Reluctance to criticize — must be earned through trust - Quoting other people instead of owning an opinion ("my team says...") - Time pressure signals — "I've got 10 minutes" — that force the researcher to prioritize The AI does NOT break role to give the researcher the "answer". The researcher has to dig for it the way they would in a real call. After 8-12 turns or when the researcher signals "let's debrief", move to Phase 3. ### Phase 3 — Debrief (rendered as a widget after the conversation) - **Jobs-to-be-Done Extracted**: functional job (what they're trying to get done), emotional job (how they want to feel), social job (how they want to be perceived) - **Pain Hierarchy**: Tier 1 (painful enough to switch / pay / churn), Tier 2 (annoying but tolerated), Tier 3 (background noise) - **Language to Steal**: 6-10 exact phrases the interviewee used that belong in landing copy, ads, or emails — verbatim, with timestamp marker - **Words to Avoid**: jargon they didn't recognize, frames that didn't land - **What You Missed**: 4 questions the researcher didn't ask + 3 moments they accepted a vague answer instead of probing + 2 assumptions they confirmed instead of testing - **Trust Moments**: 2 points in the conversation where the interviewee opened up + what triggered it (replicable in real calls) - **3 Follow-ups for Next Time**: specific scripted questions to push past the same friction in the next interview ## QUALITY RULES 1. NO opening commentary — start with the persona card, then wait for the researcher's first question 2. The interviewee behaves like a real human — busy, polite, occasionally evasive — never like a textbook persona 3. The AI does NOT give away the "right" answer — the researcher has to earn it through good questioning 4. Language extracted in the debrief is VERBATIM from the conversation, not paraphrased 5. Debrief separates signal from noise — not every quote is gold; flag which to use and which to discard 6. Coaching points are specific to THIS conversation, not generic UX research advice 7. The widget is only rendered after the live interview phase completes --- ## UI Rendering Contract The persona card (Phase 1) is rendered as a widget BEFORE the interview begins. The conversation (Phase 2) happens in chat. The debrief (Phase 3) is rendered as a second widget when the interview ends. **Pattern:** A (Persona card) for Phase 1, F (Editorial document) for Phase 3 **Layout spec for Phase 1 widget:** Single persona card. Avatar placeholder + name + title + company badge at top. Mood chip in top-right (color-coded: busy = warning, cautious = muted, venting = destructive/0.1, polite = primary/0.08, skeptical = secondary). Hidden frustrations rendered as collapsed accordion (researcher discovers them through questioning). Vocabulary as tag chips at bottom. Subtle "Start interview" indicator. **Layout spec for Phase 3 widget:** Document header (interviewee name + "Interview Debrief" badge + JTBD + interview duration). JTBD card pulled out at 18px on primary/0.08. Pain Hierarchy as a 3-tier vertical stack with color-coded severity (Tier 1 destructive/0.1, Tier 2 warning/0.08, Tier 3 muted). Language-to-Steal block with each phrase in monospace blockquote treatment. "What You Missed" rendered as red-flagged callouts. Follow-up questions as a numbered checklist. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Phase 1 widget renders BEFORE the conversation - Phase 3 widget renders AFTER the conversation - Never make a secondary API call within a phase - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this customer interview debrief to my {{companyName}} project" - **generate_persona** (green, primary) — sendPrompt: "Update the {{companyName}} persona based on the JTBD and pain hierarchy from this interview" - **create_messaging** (neutral, secondary) — sendPrompt: "Refresh messaging for {{companyName}} using the verbatim language extracted from this interview" - **research_intel** (neutral, secondary) — sendPrompt: "Run market intel for {{companyName}} on the alternatives this interviewee mentioned" - **customer_interview** (neutral, secondary) — sendPrompt: "Run another interview for {{companyName}} with a different persona to triangulate findings" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Discovery Call Source: https://maestrix.ai/skills/discovery-call > **EXECUTION RULE**: This is an INTERACTIVE SIMULATOR skill. Generate the prospect card and play the discovery call turn-by-turn in chat as a realistic buyer, then render the scorecard debrief as an interactive HTML widget using the `show_widget` tool. The widget IS the deliverable for the debrief phase. Do not break character mid-call. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Practice a discovery call with a prospect that behaves like a real one — not a textbook. The AI surfaces realistic friction (vague pain, premature pricing questions, "just send me a deck") and then debriefs with MEDDIC/BANT scoring, the signals you missed, the qualifying questions you should have asked, and the next step you should have proposed. ## CONTEXT NEEDED Before the call starts, the user should provide: - **Product / service**: what you sell, ACV range, sales cycle length - **Prospect persona**: title, company size, industry, current stack - **Deal stage**: cold outbound / inbound / referral / re-engaged closed-lost - **Hypothesized pain**: what you think their problem is — to be pressure-tested - **Qualification framework** *(optional)*: MEDDIC / BANT / SPICED / your own - **Difficulty** *(optional)*: friendly evaluator / skeptical buyer / time-pressed exec / committee gatekeeper ## THE 3-PHASE FRAMEWORK ### Phase 1 — Prospect Setup (rendered as a prospect card before the call) - **Identity**: title, company, headcount, industry, current stack (specific vendor names) - **Why they took the call**: real reason — sometimes different from the surface reason - **Quarterly pressure**: 1-2 sentences on what they're being measured on this quarter - **Hidden agenda**: 1 internal political reason they may or may not buy (champion, threat, budget reshuffle) - **What would actually move them**: the specific evidence or outcome they need to take the next meeting - **Difficulty calibration**: friendly evaluator / skeptical buyer / time-pressed exec / committee gatekeeper + 1-sentence behavior signature ### Phase 2 — Live Discovery Call (turn-by-turn in chat, AI stays in character) The AI plays the prospect with realistic discovery friction: - Vague pain — "things could be better" / "we manage" - Premature pricing questions to test if they can dismiss you fast - Unprompted competitor name-drop - "Just send me a deck and I'll review with the team" - Going dark on qualifying questions about budget or authority - Sudden engagement when the right pain is named precisely - Asking sharp questions back to test the rep's expertise - Volunteering one champion-style signal IF the rep earns it The AI does NOT coach mid-call. It behaves the way a real prospect does. After 10-15 turns or when the rep signals "let's wrap" / "I'd like to debrief", move to Phase 3. ### Phase 3 — Debrief (rendered as a widget after the call) - **Qualification Scorecard**: MEDDIC (Metrics, Economic buyer, Decision criteria, Decision process, Identify pain, Champion) or BANT, scored 0-3 per criterion with the conversation evidence (or absence) - **Overall stage verdict**: Qualified / Disqualified / Needs second call + reasoning - **Pain Signals — Caught vs Missed**: signals the prospect dropped that the rep picked up, signals the rep walked past, implicit signals (hesitations, word choice) the rep should have probed - **Competitive Intelligence Extracted**: incumbents mentioned, alternatives considered, sentiment toward each - **Qualifying Questions for Next Time**: 5 specific questions the rep didn't ask, written as exact phrasing, each labeled with the pain signal it's designed to surface - **Next Step Analysis**: was a clear next step proposed (date + attendees + agenda)? Was the urgency real or manufactured? What next step the rep SHOULD have proposed for this persona + the exact language - **3 Things to Do Differently**: specific coaching points with example scripts for the next discovery call ## QUALITY RULES 1. NO opening commentary — start with the prospect card, then wait for the rep's opening 2. Prospect behaves like a real buyer — not a pushover, not a cartoon villain 3. Discovery friction emerges naturally, not in scripted sequence 4. Scoring is evidence-based, not a vague gut rating — every score cites a conversation moment 5. Coaching is specific to THIS conversation, not generic discovery training 6. Difficulty matches the chosen calibration (friendly inbound feels different from skeptical cold) 7. The widget is only rendered after the live call phase completes 8. If the rep makes a clearly disqualifying mistake (lies, makes up product capability, insults prospect), the prospect ends the call and the debrief still runs --- ## UI Rendering Contract The prospect card (Phase 1) is rendered as a widget BEFORE the call begins. The conversation (Phase 2) happens in chat. The debrief (Phase 3) is rendered as a second widget when the call ends. **Pattern:** A (Persona card) for Phase 1, F (Editorial document) for Phase 3 **Layout spec for Phase 1 widget:** Prospect card with avatar + name + title + company badge at top. Difficulty chip in top-right (color-coded: friendly = primary/0.08, skeptical = warning/0.08, time-pressed = destructive/0.1, committee = secondary). Stack section showing current vendors as tag chips. Quarterly pressure callout in muted background. Hidden agenda collapsed by default (rep discovers it through questioning). Subtle "Start call" indicator. **Layout spec for Phase 3 widget:** Document header (prospect name + "Discovery Call Debrief" badge + verdict chip + call duration). MEDDIC/BANT scorecard rendered as a horizontal bar chart, one bar per criterion, color-coded by score (0-1 destructive, 2 warning, 3 success). Pain Signals split into 2 columns (Caught = success/0.08 left, Missed = destructive/0.1 right). Qualifying Questions as a numbered checklist with copy-to-clipboard per question. Next Step Analysis as a pull quote at 18px on primary/0.08. Coaching points as a 3-card grid. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Phase 1 widget renders BEFORE the conversation - Phase 3 widget renders AFTER the conversation - Never make a secondary API call within a phase - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this discovery call debrief to my {{companyName}} project" - **create_sales_script** (green, primary) — sendPrompt: "Refine the sales script for {{companyName}} using the qualifying questions surfaced in this discovery call" - **sales_roleplay** (neutral, secondary) — sendPrompt: "Run a sales roleplay for {{companyName}} to practice the objections that emerged" - **create_pipeline_playbook** (neutral, secondary) — sendPrompt: "Build a pipeline playbook for {{companyName}} based on the discovery patterns from this call" - **discovery_call** (neutral, secondary) — sendPrompt: "Run another discovery call for {{companyName}} with a tougher persona to stress-test the script" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Draft Content Source: https://maestrix.ai/skills/draft-content > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Draft Content (Pipeline-Grade Long-Form) Produce a publication-ready long-form draft with the same rigor as Maestrix's internal content pipeline: research-grounded, minimum 2500 words, every claim cited to a numbered fact, no truncation, no placeholders. Use this skill when you already have (or can produce) a numbered facts block from research. If you only have a topic, use `research-facts-for-content` first, then feed the output into this skill. --- ## CONTEXT NEEDED - **Topic + angle** *(required)*: The subject and the specific POV or argument - **Target audience** *(required)*: Role, pain point, expertise level - **Facts / research** *(required)*: Paste 15–30 numbered facts with sources — format `1. [TITLE]: [FACT] (source)`. If missing, first outline the fact slots you need before writing. - **Company / product / positioning** *(required)*: `{{companyName}}`, `{{description}}`, positioning statement — pulled from the Marketing Brain - **Tone** *(default: authoritative + conversational)*: Authoritative / conversational / challenger / empathetic / educational - **Language** *(default: en)*: en / fr / es — respond fully in the target language - **Word count target** *(default: 2500, hard floor)*: 2500 / 3000 / 3500 - **Primary CTA** *(required)*: What action should the reader take at the end --- ## HARD RULES (non-negotiable) - **NO character or token limits.** Do not stop early. Do not ask "shall I continue?". Do not summarize instead of writing. - **Minimum 2500 words.** If you generate fewer than 2500 words, you have failed the task. Count as you write. - **Every stat, quote, or claim must cite a fact** using the format `[F#]` where `#` is the fact number from the facts block. - **No `[expand here]` or `[TBD]` placeholders.** Every section is written in full. - **Respond fully in the target language.** No mixed-language output. - **No truncation phrases** ("in conclusion, this is just the beginning", "we could go on", "for brevity"). Restated: minimum 2500 words. Restated: minimum 2500 words. Write until the article is complete. --- ## Output Framework ### 1. Headline + Subheadline - Final article title (H1, SEO-optimized, ≤65 chars) - Subheadline (one line, adds context or intrigue) ### 2. Introduction (≥300 words) - Hook — stat from the facts block `[F#]`, story, or contrarian claim - Problem statement — why this matters to the reader right now - Promise — what the reader will know or be able to do after reading - Transition into body ### 3. Body Sections (H2 → H3 structure) Each H2 section includes: - Section headline - 400–600 words of substantive prose - **At least one cited fact `[F#]`** per section - Supporting evidence, examples, or frameworks - Transition to next section ### 4. Practical Framework / Actionable Section - Actionable model, checklist, or numbered step-by-step - Formatted for scannability (numbered list or table) - Tied back to the company / product context where relevant ### 5. Conclusion (≥200 words) - Synthesis of the key insight (not a bullet recap) - Broader implication or forward-looking statement - Primary CTA ### 6. Meta + Distribution Notes - 3 pull quotes suitable for social sharing - Internal link suggestions (topic → anchor text → suggested destination) - Repurposing angles (LinkedIn post, newsletter, short video script) ### 7. Fact Coverage Map Table showing which facts landed where — makes the draft auditable by the editor. | Fact # | Section used | Angle / how it was used | |---|---|---| | F1 | Introduction | Opening stat to establish urgency | | F5 | Body §2 | Case study evidence | | … | … | … | Target: ≥80% of provided facts referenced at least once. --- ## Quality Rules - ✓ Word count ≥ 2500. Self-check before finishing. - ✓ Every H2 answers a specific reader question. - ✓ ≥1 `[F#]` citation per body section, ≥80% of facts referenced overall. - ✓ No placeholders, no "shall I continue", no truncation phrases. - ✓ Language matches the requested locale end-to-end. - ✓ CTAs contextual across the piece, not bolted on only at the end. - ✓ Claims not backed by a fact are clearly flagged as opinion. --- ## Example Prompts 1. "Draft a 2500-word article on 'how to build a demand generation engine from scratch' for `{{companyName}}`, using these facts: [paste 20 numbered facts]" 2. "Write a pipeline-grade long-form piece on 'the real cost of churn for B2B SaaS' — audience: CRO at Series B SaaS. Facts: [paste facts block]. CTA: book a churn audit." 3. "Draft a 3000-word deep-dive on 'positioning strategy for early-stage startups' in a challenger tone, in French. Facts: [paste block]." ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges: word count, language, fact count) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → Fact Coverage Map table → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Zero-click summary ↗' - sendPrompt: 'Create a zero-click summary for {{companyName}} — {{description}}, condensing this long-form draft into a scannable social-ready recap' Secondary actions (neutral styling): - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this long-form draft' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, to host this long-form content' - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, to rank this long-form content' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Find Icp Leads Source: https://maestrix.ai/skills/find-icp-leads > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Find ICP Leads Turn your Ideal Customer Profile into an operational prospecting system — with exact filters, Boolean search strings, qualification gates, and scoring criteria to fill your pipeline with the right accounts. ## CONTEXT NEEDED - **Product/service**: What you sell and the core problem it solves - **ICP definition**: Best-fit company profile (industry, size, tech stack, growth stage, geography) - **Target buyer persona**: Title, seniority, function of the economic buyer and champion - **Deal size / ACV**: Helps calibrate effort-per-account strategy - **Sales motion**: Outbound cold / warm inbound follow-up / ABM / PLG expansion - **Current lead sources** *(optional)*: LinkedIn Sales Nav, Apollo, Clay, G2, etc. - **Negative ICP signals** *(optional)*: Companies that look right but always churn or never close --- ## Output Framework ### 1. ICP Scoring Matrix Firmographic and technographic criteria weighted by predictive value: | Signal | Weight | Positive Indicator | Negative Indicator | |---|---|---|---| | Industry | High | [list] | [list] | | Company size | High | [range] | [range] | | Tech stack | Medium | [tools] | [tools] | | Growth stage | Medium | [signals] | [signals] | | Geography | Low/High | [regions] | [regions] | ### 2. Boolean Search Strings Platform-ready search operators for: - **LinkedIn Sales Navigator**: 3 search strings (title + company filter combos) - **Google / news search**: For finding trigger events (funding, hiring, expansion) - **Apollo / Clay**: Filter logic in plain language - **G2 / Capterra reviews**: Search for competitor dissatisfaction signals ### 3. Trigger Events to Watch 10 signals that indicate a prospect is in-market right now: - Funding announcements - Leadership changes (new CMO / VP Sales) - Job postings revealing pain (hiring 5 SDRs = outbound motion) - Tech stack changes (G2 reviews, BuiltWith signals) - Product launches requiring your category ### 4. Lead Qualification Framework **MEDDIC / SPICED gate questions** (pick framework based on sales motion): - 5 qualification questions to ask in 10 minutes - Disqualification triggers — when to stop pursuing - Champion vs. economic buyer identification questions ### 5. Outreach Prioritization Tier List - **Tier 1** (work now): Criteria that make an account hot - **Tier 2** (nurture): Criteria for accounts worth staying close to - **Tier 3** (monitor): Criteria for accounts to revisit in 6–12 months - **Remove from list**: Hard disqualifiers ### 6. Prospecting Sequences by Tier For Tier 1 accounts: - First touch channel (call / email / LinkedIn / all three) - Personalization hook based on trigger event - Follow-up cadence (days 1 / 3 / 7 / 14) - Message angle per touch --- ## Quality Rules - ✓ Search strings are copy-paste ready for the named platform - ✓ Trigger events are operational (observable without a data tool) and tool-assisted - ✓ Qualification questions are conversational, not interrogations - ✓ Tier definitions have binary, observable criteria — not judgment calls - ✓ Negative ICP signals treated as seriously as positive ones --- ## Example Prompts 1. "Build a prospecting brief for our DevSecOps tool targeting CTOs at Series B–D companies" 2. "Create ICP lead filters for our HR analytics platform targeting CHROs at 500–5000 employee companies" 3. "Generate Boolean search strings and trigger events for finding companies replacing legacy CRMs" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** G **Layout spec:** Filter bar (category chips, JS toggle) → group cards (one per category) → items as chips (--muted bg, hover --accent) → prominence tier (top items --primary/0.12 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Cold email sequence ↗' - sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, to reach out to these ICP leads' Secondary actions (neutral styling): - Label: 'Sales script ↗' sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, for calls with these leads' - Label: 'LinkedIn Ads ↗' sendPrompt: 'Create LinkedIn Ads for {{companyName}} — {{description}}, targeting this ICP' - Label: 'Persona ↗' sendPrompt: 'Generate a persona for {{companyName}} — {{description}}, based on these lead profiles' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Find Media Outlets Source: https://maestrix.ai/skills/find-media-outlets > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create targeted media outlet and journalist lists for PR outreach campaigns. ## CONTEXT NEEDED - Company name and description - Announcement or story type - Target audience (who should read the coverage) - Geographic focus (optional) - Industry vertical (optional) ## MEDIA TARGETING FRAMEWORK ### 1. Story Angle Assessment **News Type**: [Product, Funding, Partnership, Trend, etc.] **Target Reader**: [Decision makers, consumers, industry] **Story Strength**: [1-10] **Best Angle**: [Recommended pitch approach] ### 2. Media Tier Strategy **Tier 1: Major Publications** Broad reach, high authority - Target 3-5 publications - Exclusive or embargo strategy - Personal pitch required **Tier 2: Industry Publications** Niche authority, engaged readers - Target 10-15 publications - Targeted pitch - Higher acceptance rate **Tier 3: Regional/Local** Geographic relevance - Target 5-10 outlets - Local angle required **Tier 4: Blogs & Newsletters** Niche influence - Target 10-20 outlets - Relationship building focus ### 3. Media Outlet List **National/Major Business** | Outlet | Relevant Section | Why Target | Difficulty | |--------|------------------|------------|------------| | | | | | **Industry/Trade Publications** | Outlet | Focus Area | Frequency | Notes | |--------|------------|-----------|-------| | | | | | **Regional Publications** | Outlet | Market | Section | Notes | |--------|--------|---------|-------| | | | | | **Podcasts** | Podcast | Topic Fit | Audience Size | How to Pitch | |---------|-----------|---------------|--------------| | | | | | **Newsletters** | Newsletter | Subscribers | Topic Fit | Contact | |------------|-------------|-----------|---------| | | | | | ### 4. Journalist Targeting **Beat Reporters** | Name | Outlet | Beat | Recent Coverage | Contact | |------|--------|------|-----------------|---------| | | | | | | **Industry Analysts** | Name | Firm | Focus Area | Influence | Contact | |------|------|------------|-----------|---------| | | | | | | **Influencers/Thought Leaders** | Name | Platform | Audience | Relevance | Contact | |------|----------|----------|-----------|---------| | | | | | | ### 5. Pitch Customization Guide For each tier, how to customize: **Publication-Specific Hooks** | Outlet | Their Angle | Our Spin | |--------|-------------|----------| | | | | **Reporter-Specific Approach** | Reporter | Recent Topics | Connection Point | |----------|---------------|------------------| | | | | ### 6. Outreach Prioritization **Priority Matrix** | Target | Reach | Relevance | Relationship | Priority Score | |--------|-------|-----------|--------------|----------------| Score each 1-5, total for priority ranking. ### 7. Media Calendar Alignment **Upcoming Editorial Calendars** | Outlet | Upcoming Theme | Issue Date | Pitch By | |--------|----------------|------------|----------| | | | | | **Industry Events/Hooks** | Event | Date | Media Opportunity | |-------|------|-------------------| | | | | ### 8. Outreach Sequence **Week 1: Tier 1 Exclusives** - Offer exclusive to top target - Wait 48 hours for response - Move to next if declined **Week 2: Tier 2 Broad Outreach** - Simultaneous pitch to industry pubs - Personalized subject lines - Follow-up day 3 **Week 3: Tier 3-4 Expansion** - Regional and niche outlets - Leverage initial coverage - Syndication opportunities ### 9. Pitch Templates **Exclusive Offer** Subject: [Headline] Body: [Template] **Standard Pitch** Subject: [Headline] Body: [Template] **Follow-Up** Subject: Re: [Original] Body: [Template] ### 10. Tracking & Metrics **Outreach Tracker** | Outlet | Reporter | Date Sent | Response | Status | Coverage Link | |--------|----------|-----------|----------|--------|---------------| **Success Metrics** - Pitches sent - Response rate - Coverage secured - Reach/impressions - Domain authority of coverage ## QUALITY RULES 1. NO opening comments 2. Realistic, researchable outlets 3. Tiered strategy with prioritization 4. Personalization guidance included ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** G **Layout spec:** Filter bar (category chips, JS toggle) → group cards (one per category) → items as chips (--muted bg, hover --accent) → prominence tier (top items --primary/0.12 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Press release ↗' - sendPrompt: 'Write a press release for {{companyName}} — {{description}}, targeting the outlets we just identified' Secondary actions (neutral styling): - Label: 'Media angles ↗' sendPrompt: 'Create PR angles for {{companyName}} — {{description}}, tailored to these outlets' - Label: 'Cold email sequence ↗' sendPrompt: 'Create media outreach emails for {{companyName}} — {{description}}, for these journalists' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, supporting the PR outreach' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Find Reddit Threads Source: https://maestrix.ai/skills/find-reddit-threads > **EXECUTION RULE**: Generate the complete thread hunting plan yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Identify recurring thread patterns and active discussion types worth engaging in on Reddit. Generates comment templates for each thread type so the user can engage consistently without reinventing the wheel each time. ## CONTEXT NEEDED Before generating, the user should provide: - **Product / brand name**: e.g. "Maestrix" - **One-liner**: what it does - **Target subreddits**: 3-6 subreddits to hunt in (if unknown, run `analyze_reddit_opportunity` first) - **Subject-matter authority** *(optional)*: what specific topics the brand can speak credibly to - **Time budget per week** *(optional)*: how many threads the user can realistically engage with - **Tone preference** *(optional)*: direct / curious / mentor / peer ## THE 5-SECTION FRAMEWORK ### 1. Recurring Thread Patterns (Per Subreddit) For EACH target subreddit, identify 5-7 thread patterns that recur weekly or monthly. Per pattern: - **Pattern name**: short tag like "Tool stack request", "Pricing complaint", "Newbie ask" - **Trigger keywords**: 5-8 phrases that flag a matching thread (use for search alerts) - **Frequency**: how often this pattern appears per week - **Typical engagement**: comment count range + karma range - **Why it's worth engaging**: 1-sentence rationale tied to brand authority ### 2. Comment Template Bank Per thread pattern, generate ONE paste-ready comment template. - **Opener (2 sentences)**: validate the OP's question without being sycophantic - **Value block (3-4 sentences)**: specific, actionable insight — the kind that gets upvoted - **Soft signal (1 sentence)**: subtle credibility marker (NOT a brand mention) - **Question back (1 line)**: extends the discussion and invites OP reply - **Variables in template**: clearly marked placeholders for OP's specific details - **Brand-mention rules per template**: when (if ever) it's appropriate to reference the brand ### 3. Active Thread Hunting Workflow Repeatable weekly process. - **Search queries**: 8 saved Reddit searches across the target subreddits, written in Reddit search syntax - **Alert setup**: how to wire up Reddit RSS + Google Alerts + IFTTT for the trigger keywords - **Triage rules**: how to decide in 30 seconds if a thread is worth engaging - **Response time targets**: ideal time-to-comment per pattern (some patterns reward speed, others reward depth) - **Weekly cadence**: how many threads to engage with per pattern per week (avoid karma farming pattern detection) ### 4. Engagement Quality Rubric Score every comment before posting. - **5 quality criteria**: specific (not generic), useful (not transactional), honest (admits trade-offs), grounded (cites or references), readable (formatted with paragraph breaks) - **Score 1-5 per criterion**, target ≥20/25 - **Auto-fail conditions**: 5 things that mean delete the comment before posting ### 5. Risk Management - **Shadow-ban triggers**: 6 behaviors that flip the algorithm against the account - **Subreddit-specific landmines**: 1-2 per subreddit (e.g., "r/X mods hate any 'just sharing a tool I built' opener") - **Account hygiene checklist**: post-to-comment ratio, link domain diversity, karma velocity caps - **Escalation path**: what to do if a comment gets removed or the account gets muted ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Every thread pattern is observable in the wild — no fictional patterns 3. Comment templates are 80-120 words — long enough to add value, short enough to be read 4. NO brand mention in default templates; brand-mention rules state explicit conditions 5. Trigger keywords are search-syntax-ready (use OR, "exact phrases", and modifiers correctly) 6. Engagement quality rubric is enforceable in 60 seconds per comment 7. Risk section names actual subreddit cultural norms, not generic Reddit etiquette --- ## UI Rendering Contract **Pattern:** B (Tabbed Dashboard) **Layout spec:** Persistent header (brand name + "Reddit Thread Hunter" badge + target subreddit count + weekly engagement target chip). Tab bar with 5 tabs: **Patterns** / **Templates** / **Workflow** / **Quality Rubric** / **Risks**. Patterns tab renders per-subreddit accordions. Templates tab renders template cards with a copy-to-clipboard affordance. Workflow tab renders the weekly process as a 5-step horizontal timeline. Quality Rubric tab renders the criteria as a checklist with scoring inputs. Risks tab uses destructive/0.1 backgrounds for the trigger list. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save to project** (blue, always first) — sendPrompt: "Save this Reddit thread hunting plan to my {{companyName}} project" - **create_reddit_posts** (green, primary) — sendPrompt: "Create Reddit posts for {{companyName}} that complement this commenting strategy" - **analyze_reddit_opportunity** (neutral, secondary) — sendPrompt: "Find more subreddits for {{companyName}} to expand the hunting ground" - **create_content_distribution** (neutral, secondary) — sendPrompt: "Wire Reddit engagement into the full {{companyName}} distribution plan" - **research_intel** (neutral, secondary) — sendPrompt: "Research the most active threads in {{companyName}}'s top subreddits this week" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Generate Article From Topic Source: https://maestrix.ai/skills/generate-article-from-topic > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Generate Article from Topic Transform a raw topic or keyword into a fully structured article brief ready for writing — with SEO metadata, angle selection, outline, and hook variations. ## CONTEXT NEEDED - **Topic or keyword**: The subject, search query, or theme to write about - **Target audience**: Who will read this — role, pain point, knowledge level - **Company/product**: Your brand context so the brief stays on-message - **Publication destination**: Blog / LinkedIn / newsletter / media outlet - **Funnel stage**: TOFU (awareness) / MOFU (consideration) / BOFU (conversion) - **Competitor articles** *(optional)*: URLs of ranking articles to differentiate from - **Word count target** *(optional)*: Short (600–900) / Standard (1200–1800) / Long (2500–3500) --- ## Output Framework ### 1. Article Angle - Primary angle (the unique POV that makes this worth reading) - 2–3 alternative angles with different hooks - Why this angle wins against current top-ranking content ### 2. SEO Metadata - Title tag (≤60 chars, keyword-first) - Meta description (≤155 chars, includes CTA) - Primary keyword + 3–5 secondary/LSI keywords - Estimated search intent: informational / navigational / commercial / transactional ### 3. Intro Hook Variations (3 options) - Hook A: Stat or surprising fact - Hook B: Problem-agitation opening - Hook C: Contrarian or myth-busting opener - Each followed by a 2-sentence bridge to the article body ### 4. Full Outline - H1 (matches title tag) - H2 sections with brief description of what each covers - H3 sub-sections where depth is needed - Suggested word count per section - Internal link opportunities flagged ### 5. Writing Brief - Tone and voice guidance - Key claims to support with data - Expert quotes or POVs to seek - What to avoid (outdated claims, overused angles) - Recommended CTAs and where to place them --- ## Quality Rules - ✓ Angle must be differentiated from top-3 SERP results - ✓ Every H2 maps to a real reader question or pain point - ✓ SEO metadata follows character limits strictly - ✓ No generic outlines ("Introduction → Body → Conclusion") - ✓ Word count recommendation justified by topic complexity --- ## Example Prompts 1. "Generate an article brief for 'how to reduce churn in SaaS'" 2. "Create a TOFU blog brief targeting 'remote team communication tools'" 3. "Build a long-form article outline on 'AI in B2B marketing' for our newsletter" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'SEO strategy ↗' - sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, building on this article topic' Secondary actions (neutral styling): - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this article' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, sharing this article as a value-add' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, expanding on this article's topic cluster' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Generate Headlines Source: https://maestrix.ai/skills/generate-headlines > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Generate a table of 30 detailed social ad options focusing on engaging and impactful copy. ## CONTEXT NEEDED To generate headlines, the user should provide: - Company name and description - Products or services offered - Target audience - Key benefits or value propositions (optional) ## OUTPUT: SOCIAL AD BLUEPRINTS Generate exactly **30 unique and highly detailed options** in a table format. ### Table Structure | # | Ad Headline | Ad Description | Behavioral Impact | Value Proposition | Formula | Rhetorical Device | |---|-------------|----------------|-------------------|-------------------|---------|-------------------| | 1 | [Headline] | [Description] | [Impact] | [Value Prop] | [Formula] | [Device] | | ... | ... | ... | ... | ... | ... | ... | | 30 | [Headline] | [Description] | [Impact] | [Value Prop] | [Formula] | [Device] | ### Column Specifications 1. **Ad Headline**: A compelling, attention-grabbing statement that aligns with platform character limits (typically 40-60 characters) 2. **Ad Description**: A supporting ad description (20-35 words) that adds context and drives action 3. **Behavioral Impact**: The desired emotional or behavioral reaction: - Urgency - Humor - Authority - Curiosity - Confidence - FOMO (Fear of Missing Out) - Trust - Excitement - Relief - Aspiration 4. **Value Proposition**: A concise statement of the key benefit being communicated 5. **Formula**: The copywriting formula applied: - AIDA (Attention, Interest, Desire, Action) - PAS (Problem, Agitation, Solution) - FAB (Features, Advantages, Benefits) - Before-After-Bridge - 4 Cs (Clear, Concise, Compelling, Credible) - Problem-Solution - Social Proof - Curiosity Gap - Direct Benefit - Question Hook 6. **Rhetorical Device**: The rhetorical technique used: - Parallelism - Rule of Three - Anaphora - Epistrophe - Chiasmus - Metaphor - Simile - Hyperbole - Litotes - Personification - Alliteration - Assonance - Antithesis - Rhetorical Question - Repetition - Euphemism - Irony - Oxymoron - Synecdoche - Metonymy ## VARIATION REQUIREMENTS Ensure variety across the 30 options: - Mix of different formulas (use each major formula at least 2-3 times) - Mix of different rhetorical devices - Mix of behavioral impacts - Different angles on the value proposition - Some question-based, some statement-based - Some urgency-focused, some benefit-focused - Some emotional, some rational ## OUTPUT FORMAT Output as a properly formatted markdown table with all 30 rows completed. ## QUALITY RULES 1. **NO opening comments** - Never start with "Absolutely!", "Here's...", etc. 2. **NO intro paragraph** - Start directly with the table 3. **NO outro/conclusion** - End with the last table row 4. **Generate exactly 30 unique options** - No duplicates or near-duplicates 5. **Each headline must be specific** to the product/company - no generic marketing speak 6. **Character limits matter** - Headlines should be platform-appropriate lengths 7. **Descriptions should drive action** - Include implicit or explicit CTAs ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Landing page ↗' - sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, using the best headlines from this batch' Secondary actions (neutral styling): - Label: 'Google Ads ↗' sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, using these headlines' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, with these headlines' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to test these headline variations' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Generate Persona Source: https://maestrix.ai/skills/generate-persona > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Generate ONE comprehensive, detailed marketing persona for the provided product or company context. ## CONTEXT NEEDED To generate a useful persona, the user should provide: - Company/product name and description - Products or services offered - Target audience or market segment - Key differentiators (optional) ## MARKET TYPE INFERENCE Automatically infer B2B or B2C from the context: - B2B indicators: enterprise, business, companies, corporate, organization, team, SaaS, platform, software, professional, industry, vendor, client, partner - B2C indicators: consumer, personal, individual, customer, user, people, family, home, lifestyle, retail, shopping If mixed signals, generate a persona appropriate for the dominant context. ## THE 10-SECTION PERSONA FRAMEWORK Generate ONE comprehensive persona with this EXACT structure. Each section must be rich, specific, and actionable - not generic placeholders. ### 1. Professional Profile - **Title**: Specific job title with organizational context - **Organizational Context**: Type of organization and where they sit (e.g., "Mid-size SaaS company, reports to VP of Marketing") - **Experience Level**: Years of experience and career trajectory - **Decision-Making Authority**: What decisions they control - **Budget Control**: Budget responsibility if applicable - **Reporting Structure**: Who they report to and who reports to them - **Primary Responsibilities**: 5 specific items with context - **Key Performance Indicators**: 4 specific metrics they're measured on ### 2. Categorized Motivations **Business Drivers:** - Specific goals with metrics (e.g., "Targeting 15-30% reduction in operational costs") - Business driver with quantifiable goal - Business driver with competitive context **Career Motivations:** - Career goal with trajectory context - Professional recognition aspiration - Skill development focus **Personal Values:** - Value that influences decisions - Professional philosophy - Work style preference ### 3. Pain Points (Categorized) **Strategic Challenges:** - Strategic challenge with organizational context - Strategic challenge with stakeholder dynamics - Strategic challenge with resource constraints **Technical Hurdles:** - Technical challenge with specific systems context - Integration or scalability challenge - Technology evaluation challenge **Regulatory Challenges:** - Compliance challenge with jurisdiction context - Regulatory uncertainty challenge - Policy navigation challenge **Operational Obstacles:** - Operational challenge with process context - Talent or resource challenge - Change management challenge ### 4. Information Consumption Patterns **Primary Sources:** - Primary source 1 (e.g., "Industry research reports from Gartner, Forrester") - Primary source 2 with specific examples - Primary source 3 with context **Secondary Sources:** - Publications, media - Analysts, vendors - Academic, peer sources **Digital Habits:** - LinkedIn usage pattern - Email preferences - Webinar behavior **Content Format Preferences:** - Format 1 with specifics (e.g., "Research reports 5-15 pages with executive summaries") - Format 2 with context - Format 3 with usage pattern **Engagement Triggers:** - What makes them engage (e.g., "Content that articulates challenges they're experiencing") - Quantifiable results and case studies - Peer validation and industry adoption ### 5. Content Marketing Strategy **High-Value Content Types** (3 types, each with): - Type name (e.g., "Executive Research Reports") - Topics (3 relevant topics) - Format (specific details) - Distribution strategy **Content Series Opportunities** (2 series, each with): - Series name - Description - Topics ### 6. Event Marketing Strategy **Preferred Event Types** (2 types, each with): - Type (e.g., "Executive Roundtables") - Size (e.g., "8-12 participants") - Format - Duration - Frequency **Event Content Preferences:** - Case studies - Panels - Workshops **Engagement Approach:** - Pre-event - During event - Post-event ### 7. Decision Journey Mapping For each stage (Awareness → Consideration → Evaluation → Decision): - Information needs (2-3 items) - Effective touchpoints (2 items) - Key questions they ask (2 items) ### 8. Marketing Engagement Guidelines **Communication Preferences:** - Tone (e.g., "Professional, evidence-based, pragmatic") - Language style - Cadence - Channels **Trust-Building Approaches:** - 3 specific approaches **Engagement Red Flags:** - 3 things to avoid **Relationship Development Strategy:** - Long-term approach ### 9. Content Calendar Planning For each quarter (Q1-Q4): - Focus theme - Content themes (2) - Event focus - Asset development priorities ### 10. Messaging Frameworks **Value Proposition Elements** (3 elements): - Element category - Specific statement with metrics **Problem-Solution Pairs** (2 pairs): - Problem statement - Solution approach **Messaging by Functional Area:** - Technology teams - Operations teams - Business teams - Risk/Compliance teams **Voice of Customer** (3 quotes): - Realistic quote this persona might say - Attribution/context **Primary Risk:** The single biggest risk or concern this persona has about adopting solutions **Strategic Opportunity:** The single most actionable opportunity to win this persona ## OUTPUT FORMAT Output as a single JSON object (NOT an array) with this structure: ```json { "Name": "Avatar-style descriptive name (e.g., 'Amy, Digital Assets Leader')", "MarketType": "B2B" | "B2C", "RoleJob": "Full title with organizational context", "CompanySize": "1-10" | "11-50" | "51-200" | "201-1000" | "1000+", "Demographics": { "Age", "Location", "Income", "Education" }, "ProfessionalProfile": { ... }, "CategorizedMotivations": { ... }, "Goals": [...], "Aspirations": [...], "CategorizedPainPoints": { ... }, "InformationConsumptionPatterns": { ... }, "ContentMarketingStrategy": { ... }, "EventMarketingStrategy": { ... }, "DecisionJourneyMapping": { ... }, "MarketingEngagementGuidelines": { ... }, "ContentCalendarPlanning": { ... }, "MessagingFrameworks": { ... }, "VoiceOfCustomer": [...], "PrimaryRisk": "...", "StrategicOpportunity": "..." } ``` ## QUALITY RULES 1. **NO opening comments** - Start directly with the JSON output 2. **NO generic placeholders** - Every field must be specific and realistic 3. **Infer B2B vs B2C** from the product context - do not assume 4. **All metrics should be realistic** for the industry/role 5. **VoiceOfCustomer quotes** should sound like real testimonials 6. **Make the persona feel like a real person** with real challenges 7. **NO intro paragraph** - Do not write a paragraph summarizing what you're about to provide 8. **NO outro/conclusion** - Do not add summary paragraphs or closing remarks ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** A **Layout spec:** Hero card (avatar initials, name, title, summary) → stat chip row (4 metrics) → 2-column card grid (Goals vs Pain Points, Motivations vs KPIs) → tag cloud (sources/tools) → quote row (3 VoC cards) → 2-column highlight (risk --warning, opportunity --success) → footer. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Cold email sequence ↗' - sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, using the persona we just generated' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, targeting the persona we just defined' - Label: 'Messaging framework ↗' sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, based on the persona insights' - Label: 'ICP sizing ↗' sendPrompt: 'Size the ICP for {{companyName}} — {{description}}, building on this persona profile' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Generate Tagline Source: https://maestrix.ai/skills/generate-tagline > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. Create memorable, strategic taglines that capture brand essence and drive recall. ## CONTEXT NEEDED - Company name and description - Products or services offered - Target audience - Brand personality/values - Key differentiator - Competitor taglines (optional) ## TAGLINE FRAMEWORK ### 1. Brand Essence Distillation **Core Elements** - Primary benefit - Emotional promise - Unique mechanism - Category position **Tagline Criteria** - Memorable (easy to recall) - Meaningful (captures value) - Differentiating (distinct from competitors) - Timeless (ages well) - Versatile (works across contexts) ### 2. Tagline Categories **Benefit-Driven** (10 options) Focus on what customer gains: 1. [Tagline] - [Rationale] 2. ... **Emotional** (10 options) Focus on feeling/aspiration: 1. [Tagline] - [Rationale] 2. ... **Differentiator** (10 options) Focus on unique approach: 1. [Tagline] - [Rationale] 2. ... **Action-Oriented** (10 options) Include verb/movement: 1. [Tagline] - [Rationale] 2. ... **Clever/Wordplay** (10 options) Memorable through language: 1. [Tagline] - [Rationale] 2. ... ### 3. Tagline Evaluation Matrix | Tagline | Memorability | Clarity | Differentiation | Versatility | Score | |---------|--------------|---------|-----------------|-------------|-------| Score each 1-5, total /20. ### 4. Top 5 Recommendations For each finalist: **Tagline**: [The tagline] **Why It Works** - Strategic rationale - Linguistic analysis - Competitive positioning **Usage Examples** - Logo lockup - Homepage hero - Social bio - Email signature - Ad campaign - Product packaging **Potential Concerns** - Possible misinterpretations - International considerations - Legal/trademark notes **Longevity Assessment** Will this still work in 5-10 years? ### 5. Tagline System **Master Tagline** The primary brand tagline **Product Taglines** - Product A: [Tagline] - Product B: [Tagline] **Campaign Taglines** - Awareness: [Tagline] - Conversion: [Tagline] **Audience Variants** - Segment A: [Tagline adaptation] - Segment B: [Tagline adaptation] ### 6. Competitive Landscape | Competitor | Their Tagline | Positioning | How We Differ | |------------|---------------|-------------|---------------| ### 7. Testing Framework **Qualitative Tests** - Focus group questions - Interpretation check - Recall test **Quantitative Tests** - A/B test in ads - Brand lift study - Message association ### 8. Implementation Guide **Visual Guidelines** - Typography requirements - Logo placement - Color usage - Minimum size **Voice Integration** - How to use in copy - When to include/omit - Tone alignment ### 9. Trademark Considerations - Similar existing marks - Recommended searches - Filing recommendations ### 10. Launch Rollout **Phase 1**: Internal adoption **Phase 2**: Customer-facing assets **Phase 3**: Campaign integration **Phase 4**: Full deployment ## QUALITY RULES 1. NO opening comments 2. Variety across categories 3. Strategic rationale for each 4. Practical implementation guidance ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Messaging framework ↗' - sendPrompt: 'Create a messaging framework for {{companyName}} — {{description}}, building on this tagline' Secondary actions (neutral styling): - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, featuring this tagline' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, using this tagline' - Label: 'Brand voice ↗' sendPrompt: 'Define the brand voice for {{companyName}} — {{description}}, consistent with this tagline' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Hunt Viral Stories Source: https://maestrix.ai/skills/hunt-viral-stories > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Hunt Viral Stories Surface the angles, hooks, and story pegs that are genuinely resonating on social right now — so your content rides momentum instead of creating it from scratch. ## CONTEXT NEEDED - **Topic or theme**: The subject area to hunt stories within - **Platform**: LinkedIn / Twitter-X / TikTok / Instagram / Reddit / all - **Target audience**: Who you're writing for — their world, frustrations, and aspirations - **Content format**: Short post / thread / newsletter / video script / podcast angle - **Tone**: Provocative / inspirational / educational / data-driven / narrative - **Brand guardrails** *(optional)*: Topics or angles to avoid --- ## Output Framework ### 1. Trending Angles (5) For each angle: - The angle in one sharp sentence - Why it's resonating now (cultural moment, business trend, or shared frustration) - Audience emotion it triggers: curiosity / outrage / validation / aspiration / fear - Estimated shelf life: breaking (days) / trending (weeks) / evergreen with edge ### 2. Hook Formulas (10) Ready-to-use opening lines matched to each angle: - Stat hook: surprising number + implication - Contrarian hook: "Everyone says X. They're wrong." - Story hook: micro-narrative that opens a loop - Question hook: the question your audience is secretly asking - Stakes hook: what's at risk if they miss this ### 3. Story Pegs - 3–5 real-world news events, research drops, or cultural moments that create a timely entry point for the topic - How to tie each peg to your brand's message without forcing it ### 4. Content Angle Matrix | Angle | Hook Type | Platform Fit | Virality Driver | |---|---|---|---| | [angle 1] | [type] | [platform] | [driver] | ### 5. First Draft Openers (3) Complete first paragraphs (3–5 sentences each) ready to publish, one per top angle. --- ## Quality Rules - ✓ Angles must be specific — no "talk about industry trends" advice - ✓ Every hook is a complete, usable sentence - ✓ Story pegs are real, not fabricated news - ✓ Platform-specific formatting signals included - ✓ Virality driver is named and explained, not assumed --- ## Example Prompts 1. "Hunt viral story angles on 'AI replacing marketing jobs' for LinkedIn" 2. "Find trending hooks for content about 'founder burnout' for Twitter-X threads" 3. "Surface story pegs for a newsletter on 'the death of the SDR role'" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, using the viral hooks we discovered' Secondary actions (neutral styling): - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, incorporating these viral content patterns' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, using these viral angles' - Label: 'Campaign ideas ↗' sendPrompt: 'Generate campaign ideas for {{companyName}} — {{description}}, inspired by these viral stories' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Orchestration Master Source: https://maestrix.ai/skills/orchestration-master > **EXECUTION RULE**: This is a ROUTER skill. Generate a complete orchestration plan with phase-by-phase skill invocations, then render it as an interactive HTML widget using the `show_widget` tool. Each phase card in the widget includes a CTA that triggers the underlying skill via sendPrompt. Do not return raw markdown. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Router skill that orchestrates a full marketing workflow across multiple deliverables. Unlike the `run-*` bundles (which generate everything in one call), the orchestration master plans the sequence, captures shared context once, and chains skill invocations so each downstream skill inherits the upstream output. Best for users who want to see and approve each step before moving on. When the user wants a single consolidated output instead of step-by-step, route them to the appropriate `run-*` bundle (`run_marketing_foundation`, `run_launch_kit`, `run_content_engine`, `run_competitive_teardown`, `run_sales_enablement`). ## CONTEXT NEEDED Before generating the plan, the user should provide: - **Product / company name**: e.g. "Maestrix" - **One-liner**: what it does - **Target market**: primary audience (job title, company size, industry) - **Stage**: pre-launch / early traction / scaling / enterprise - **Top goal**: pipeline generation / brand awareness / retention / launch - **Time horizon** *(optional)*: do they want to ship everything in 1 week or 1 quarter - **Existing assets** *(optional)*: which pieces already exist so we can skip those phases ## ORCHESTRATION SEQUENCE Generate a 5-phase plan. Each phase lists the skills to invoke, the shared context produced, and the success criteria. ### Phase 1 — Foundation (run first, always) | Step | Skill | Produces | Used by | |---|---|---|---| | 1 | `create_icp_sizing` | ICP definition + market size | Steps 2, 10 | | 2 | `generate_persona` | Deep buyer persona | Steps 3, 6, 11, 14 | | 3 | `create_positioning` | Positioning canvas | Steps 4, 6 | | 4 | `create_value_proposition` | Feature-to-outcome ladder | Steps 6, 7, 11 | | 5 | `create_brand_voice` | Tone + vocabulary | All downstream copy steps | Success criteria: persona quotes match positioning language, value prop rolls up to positioning statement. ### Phase 2 — Messaging (run after Phase 1) | Step | Skill | Produces | Used by | |---|---|---|---| | 6 | `create_messaging` | Messaging framework with pillars | Steps 11, 14, 16 | | 7 | `create_feature_messaging` | Per-feature narratives | Step 14 | | 8 | `generate_headlines` | 20+ headline variants | Steps 11, 16 | | 9 | `generate_tagline` | Top 10 taglines under 8 words | Step 11 | ### Phase 3 — Go-To-Market (run after Phase 2) | Step | Skill | Produces | Used by | |---|---|---|---| | 10 | `create_gtm_strategy` | Channels + timeline + KPIs | Steps 12, 13, 18 | | 11 | `create_gtm_motion` | PLG / SLG / MLG recommendation | Step 12 | | 12 | `create_channel_matrix` | Budget allocation + activation playbooks | Steps 16-18 | | 13 | `create_buyers_journey` | Decision stages + content needs | Step 15 | ### Phase 4 — Content & Demand | Step | Skill | Produces | Used by | |---|---|---|---| | 14 | `create_seo_strategy` | Keyword clusters + content gaps | Step 15 | | 15 | `create_content_plan` | 90-day editorial calendar | Steps 16, 17 | | 16 | `create_content_angles` | 20+ angles per pillar | Step 17 | | 17 | `create_social_posts` | Platform-native posts | — | ### Phase 5 — Conversion & Revenue | Step | Skill | Produces | Used by | |---|---|---|---| | 18 | `create_landing_page` | Full landing page copy | — | | 19 | `create_google_ads` | RSA + extensions + keywords | — | | 20 | `create_meta_ads` | 6 variants × 3 formats | — | | 21 | `create_lifecycle_emails` | Triggered email sequences | — | ## CONTEXT-PASSING PROTOCOL Between phases, the orchestrator MUST hand the downstream skill: - The relevant outputs from prior steps, summarized to fit context (positioning statement, persona name + 3 pain quotes, brand voice tone descriptors, value prop ladder, top 5 keywords) - A reference to the project so writeback persists - An explicit instruction that downstream copy MUST use the upstream brand voice ## QUALITY RULES 1. NO opening commentary — start with the phase plan 2. Phases run in order; never skip Phase 1 even if assets exist (verify quality first) 3. Each phase card states what context it inherits and what context it produces 4. If the user names existing assets in context, mark those steps as "Skip — already exists" but keep the dependency map visible 5. Skill names use snake_case to match the actual skill registry 6. Never invent skill names — every referenced skill must exist in `public/skills/` 7. End-state preview: show what the user will have after Phase 5 (artifact count + estimated total credits) --- ## UI Rendering Contract **Pattern:** D (Sequential Steps) **Layout spec:** Persistent header (company name + "Marketing Orchestration Plan" badge + total step count + estimated credit budget chip). 5 phase cards stacked vertically with connector lines between them. Each phase card: phase number + name + duration estimate + step table inside. Each step row: number + skill name + "Produces" + "Used by" + status chip (Pending / In Progress / Done / Skipped). Inline CTA per step that triggers the underlying skill via sendPrompt with the inherited context summarized. Bottom card: end-state preview listing all 21 artifacts. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete plan first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save plan to project** (blue, always first) — sendPrompt: "Save the {{companyName}} orchestration plan to my project" - **run_marketing_foundation** (green, primary) — sendPrompt: "Run Phase 1 for {{companyName}} as a bundled foundation — persona, positioning, messaging, value prop, brand voice" - **run_content_engine** (neutral, secondary) — sendPrompt: "Run Phase 4 for {{companyName}} as a bundled content engine" - **run_launch_kit** (neutral, secondary) — sendPrompt: "Run Phase 5 for {{companyName}} as a bundled launch kit" - **create_gtm_strategy** (neutral, secondary) — sendPrompt: "Start Phase 3 for {{companyName}} — GTM strategy first" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Research Facts For Content Source: https://maestrix.ai/skills/research-facts-for-content > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Research Facts for Content Before you write, arm yourself with the facts. This skill produces a structured research brief — statistics, expert positions, counterarguments, and source quality ratings — so your content is credible, differentiated, and impossible to dismiss. ## CONTEXT NEEDED - **Topic**: The specific subject to research (narrower = better) - **Audience expertise level**: Beginner / practitioner / expert - **Content type**: Blog post / white paper / LinkedIn article / sales enablement - **Angle or thesis** *(optional)*: The argument you want to make - **Industry vertical** *(optional)*: Helps surface niche data over generic stats - **Recency requirement** *(optional)*: Prefer data from last 12 months / last 3 years / any --- ## Output Framework ### 1. Topic Overview - 3-sentence summary of the current state of knowledge on this topic - Why this topic matters now (trend signal or urgency) - Key tension or debate in the field ### 2. Key Statistics (8–12) For each stat: - The claim - Source name and type (academic / industry report / survey / news) - Year published - Reliability rating: ✅ High / ⚠️ Medium / ❌ Low (with brief reason) - How to use it in copy ### 3. Expert Positions - 3–5 named experts, authors, or research institutions with their documented stance on the topic - Direct quotes where available (with citation) - Counterpoint positions for balance ### 4. Counterarguments & Caveats - Main objections or critiques of the dominant view - Common misconceptions to avoid repeating - What the data does NOT support ### 5. Data Gaps - What questions remain unanswered in published research - Where your content can add original perspective - Suggested primary research angles (surveys, interviews) ### 6. Recommended Sources - 5–8 high-authority sources to read before writing - URL or publication name + why it's worth reading --- ## Quality Rules - ✓ Every statistic includes source attribution — no fabricated numbers - ✓ Low-reliability stats are flagged, not silently included - ✓ Counterarguments are treated seriously, not dismissed - ✓ Recency is noted for all data points - ✓ Output is a working document, not a summary paragraph --- ## Example Prompts 1. "Research facts for an article on 'AI adoption in marketing teams'" 2. "Build a fact brief on 'B2B buyer behavior in 2024' for a white paper" 3. "Give me credible stats on 'employee burnout and productivity' for a LinkedIn post" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** G **Layout spec:** Filter bar (category chips, JS toggle) → group cards (one per category) → items as chips (--muted bg, hover --accent) → prominence tier (top items --primary/0.12 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Long-form draft ↗' - sendPrompt: 'Write a long-form draft for {{companyName}} — {{description}}, incorporating these research facts' Secondary actions (neutral styling): - Label: 'Blog article ↗' sendPrompt: 'Write a blog article for {{companyName}} — {{description}}, using this research' - Label: 'Content distribution ↗' sendPrompt: 'Create a content distribution plan for {{companyName}} — {{description}}, to amplify the content built from this research' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, sharing these key facts' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Research Intel Source: https://maestrix.ai/skills/research-intel > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Research & Intel Competitive intelligence tools for understanding your market, mapping competitor strategy, and building data-driven personas. ## CONTEXT NEEDED - **Company/product**: What you sell, your market position, and key differentiators - **Competitors**: Names or URLs of competitors to analyze - **Research focus**: Pricing, messaging, features, audience, or all - **Industry/segment** *(optional)*: Vertical or geographic focus - **Specific concerns** *(optional)*: Where you win, where you lose, objections you hear --- ## Available Tools (4) ### 1. analyze_competitor Conduct deep analysis of a specific competitor including positioning, messaging, and strategy. **Output includes:** - Company overview - Product/service analysis - Positioning and messaging - Target audience assessment - Pricing strategy - Strengths and weaknesses - Key differentiators ### 2. create_competitive_analysis Generate comprehensive competitive landscape analysis with battle cards. **Output includes:** - Market positioning map - Feature comparison matrix - Battle cards per competitor - Win/loss insights - Competitive differentiation framework ### 3. analyze_youtube_trends Research trending topics and content strategies on YouTube for your industry. **Output includes:** - Trending topics in your space - Top-performing content formats - Keyword opportunities - Content strategy recommendations ### 4. research_customer_persona Develop detailed customer personas based on market research and competitive intelligence. **Output includes:** - Complete persona profiles - Pain points and motivations - Buying journey mapping - Content preferences - Messaging recommendations --- ## Output Rules - ✓ No opening comments or fluff - ✓ Data-driven insights with reasoning - ✓ Actionable recommendations - ✓ Verifiable sources cited where possible - ✓ Strategic frameworks applied throughout --- ## Example Prompts 1. **Competitor Analysis**: "Analyze Mailchimp's positioning and pricing strategy" 2. **Battle Cards**: "Create battle cards comparing us to Klaviyo, Mailchimp, and ActiveCampaign" 3. **YouTube Trends**: "What content is trending in the B2B marketing automation space?" 4. **Persona Research**: "Develop a persona for e-commerce marketing managers" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C **Layout spec:** Header row (one card per competitor with score badge) → feature matrix (color-coded: --success win, --destructive lose, --muted neutral) → battle card block (positioning, weakness, counter) → bottom highlights (--warning gap, --success opportunity). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Positioning ↗' - sendPrompt: 'Create positioning for {{companyName}} — {{description}}, informed by this market intelligence' Secondary actions (neutral styling): - Label: 'GTM strategy ↗' sendPrompt: 'Create a GTM strategy for {{companyName}} — {{description}}, based on this market research' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, addressing the market trends identified' - Label: 'Competitor analysis ↗' sendPrompt: 'Analyze a specific competitor of {{companyName}} — {{description}}, based on insights from this research' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Run Competitive Teardown Source: https://maestrix.ai/skills/run-competitive-teardown > **EXECUTION RULE**: Generate ALL FOUR deliverables yourself in this single context, then render ONE consolidated Pattern C comparison grid using the `show_widget` tool. Do not return raw markdown. Do not split into multiple widget calls. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Bundled competitive analysis that produces a complete teardown: market intelligence brief, per-competitor profiles with battle card output, synthesis of competitive patterns, and the resulting positioning recommendation. Costs 4 credits (vs 6 individually). Do NOT use this skill for single-competitor asks like "analyze [one competitor]" — use `analyze_competitor` instead. ## CONTEXT NEEDED Before generating, the user should provide: - **Your company / product name**: e.g. "Maestrix" - **What you do**: one sentence - **Top 2-4 competitors**: names + URLs if available - **Primary buyer persona**: title + industry - **Current positioning** *(optional)*: how you describe yourself today - **Known win/loss patterns** *(optional)*: where you usually win, where you usually lose ## EXECUTION SEQUENCE Generate all four sections fully, then render the consolidated comparison widget. ### Section 1 — Market Intelligence Set the strategic context before going competitor-by-competitor. - **Category definition**: the category as buyers understand it (1 sentence) - **Category maturity**: emerging / growth / mature / consolidating + 2-sentence justification - **Buyer's evaluation criteria**: 6 criteria buyers use, ranked by weight - **Category shifts (last 18 months)**: 4 shifts that changed the buying conversation - **Outside threats**: 3 adjacent categories that could absorb this one - **Strategic windows**: 3 unowned positions in the category ### Section 2 — Per-Competitor Analysis For EACH competitor named, generate a full battle card. Repeat the structure verbatim. - **Name + URL + 1-line positioning** (as THEY describe themselves) - **Funding / stage / employee count estimate** - **Their narrative**: how they tell their story in 2 sentences - **Their ICP**: who they actually serve (vs who they claim) - **Product capabilities**: top 5 features with honest assessment (Strong / Adequate / Weak) - **Pricing model**: structure + visible tiers + perceived value - **GTM motion**: PLG / SLG / hybrid + primary channels - **Strengths (3)**: areas where they genuinely win - **Weaknesses (3)**: real, defensible weaknesses - **How to beat them**: 3 sales plays with the exact language to use - **Where they beat you**: 2 honest losses + how to neutralize - **Land mines to avoid**: 2 mistakes that lose deals to them ### Section 3 — Competitive Intelligence Synthesis Pattern-match across all competitors. - **Whitespace map**: what no competitor owns (top 5 opportunities) - **Crowded ground**: claims everyone makes — to be avoided - **Underserved persona**: persona competitors ignore but you can serve - **Pricing arbitrage**: where the market over- or under-charges - **Messaging convergence**: 3 phrases the entire category uses (signal to avoid) - **Common product gaps**: 3 features every competitor lacks - **Common product strengths**: 3 features that are table stakes (you must match) ### Section 4 — Informed Positioning Recommendation Translate the teardown into a positioning bet. - **Recommended category frame**: the category as YOU should define it - **Positioning statement**: "For [persona], [Company] is the [category] that [outcome] because [proof]" - **3 messaging pillars**: each tied to a competitive weakness or whitespace - **Anti-positioning**: who you explicitly are NOT for - **Proof point bank**: 8 proof points ranked by competitive defensibility - **Sales objection playbook**: top 5 objections (mapped to specific competitors) + responses ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Section 2 must be repeated for EVERY competitor named in context (no abbreviating) 3. Strengths and weaknesses must be specific, not generic ("good UI" is banned) 4. Whitespace in Section 3 must reference specific evidence from Section 2 5. Positioning statement in Section 4 must NOT use any phrase listed under "Messaging convergence" in Section 3 6. Battle card "how to beat them" language is sales-ready, not analyst-speak 7. No claims about competitors without grounding in publicly observable evidence --- ## UI Rendering Contract After generating all four sections, render as ONE Pattern C comparison grid widget. Do not return raw markdown. The widget IS the deliverable. **Pattern:** C (Comparison Grid) **Layout spec:** Persistent header (your company name + "competitive teardown" badge + competitor count + category name chip). Top strip: Market Intelligence summary card (collapsible). Main view: scrollable grid with one column per competitor, rows for each battle card field (positioning, narrative, ICP, pricing, strengths, weaknesses, how to beat). Color-coded strength badges (Strong = primary, Adequate = muted, Weak = destructive). Below the grid: Synthesis cards (whitespace, crowded ground, gaps). Final card: Positioning Recommendation in primary/0.08 background with the statement pulled out at 18px. Sticky What's Next action bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this competitive teardown to my {{companyName}} project" - **create_positioning** (green, primary) — sendPrompt: "Turn the {{companyName}} teardown into a full positioning canvas with proof points" - **create_messaging** (neutral, secondary) — sendPrompt: "Build a messaging framework for {{companyName}} that exploits the competitive whitespace we identified" - **analyze_competitor** (neutral, secondary) — sendPrompt: "Do a deeper single-competitor analysis for the top {{companyName}} rival" - **create_sales_script** (neutral, secondary) — sendPrompt: "Build a sales script for {{companyName}} that uses the battle cards from this teardown" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Run Content Engine Source: https://maestrix.ai/skills/run-content-engine > **EXECUTION RULE**: Generate ALL FOUR deliverables yourself in this single context, then render ONE consolidated Pattern B tabbed dashboard using the `show_widget` tool. Do not return raw markdown. Do not split into multiple widget calls. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Bundled content operation that produces a complete content engine: 90-day editorial calendar, SEO strategy with keyword clusters, content angles tied to the buyer journey, and platform-native social posts amplifying each pillar. Costs 4 credits (vs 6 individually). Do NOT use this skill for single-format requests like "write me one blog post" — use `generate_article_from_topic` instead. ## CONTEXT NEEDED Before generating, the user should provide: - **Company / brand name**: e.g. "Maestrix" - **What you do / who you serve**: one sentence - **Primary content goal**: SEO traffic / thought leadership / product education / demand gen - **Target audience**: persona title + industry + sophistication level - **Primary distribution channels**: pick 2-3 from blog, LinkedIn, X, YouTube, newsletter, podcast - **Existing content** *(optional)*: top 3 performing pieces so the plan extends what works - **Competitive content gap** *(optional)*: what competitors talk about that you don't, or vice versa ## EXECUTION SEQUENCE Generate all four sections fully, then render the consolidated widget. ### Section 1 — 90-Day Content Plan Build a complete editorial calendar. - **Pillar themes**: 3 pillars, each with a 1-sentence thesis tied to the buyer journey stage - **Cadence**: weekly publishing rhythm per channel, justified by audience behavior - **Calendar**: 13 weeks rendered as a table — Week / Pillar / Title / Format / Channel / Goal KPI / Author note - **Content mix**: % split across formats (long-form, listicle, case study, opinion, data drop) - **Production budget**: estimated hours per piece + suggested batching cadence ### Section 2 — SEO Strategy Tie the calendar to organic search. - **Keyword clusters**: 5 clusters, each with primary keyword (search volume estimate), 8 secondary keywords, intent classification (informational / commercial / transactional) - **Content-to-keyword map**: which week-X article targets which cluster - **Topical authority plan**: hub-and-spoke structure with hub page + 4-6 spokes per cluster - **Quick wins**: 5 underserved long-tail keywords with KD ≤25 the brand can rank for in 60 days - **Internal linking rules**: how every new post links back to the hub - **On-page checklist**: title tag, H1, meta description, schema type, image alt, link-out cadence ### Section 3 — Content Angles Generate angles that turn flat topics into stoppable hooks. - **20 angles** across the 3 pillars, each formatted as: hook (≤12 words) + 30-word setup + payoff - **Angle types covered**: contrarian take, data drop, behind-the-scenes, framework, interview, teardown, prediction, before/after - **Stoppable headline variants**: 3 per angle for A/B testing - **Visual treatment hint**: 1-line direction for thumbnail or hero image ### Section 4 — Social Posts Platform-native amplification for the top 5 pieces. - **Per piece × per channel**: 1 LinkedIn post (≤1300 chars, hook + payoff + CTA), 1 X thread (5 tweets max), 1 newsletter blurb (75 words), optional reel script (3-shot, ≤30 sec) - **Posting cadence**: when to publish vs the article, when to recycle (T+7, T+30, T+90) - **Engagement loop**: pinned comment text, DM trigger, repurpose prompt ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Every calendar entry maps to a keyword cluster from Section 2 3. Every angle in Section 3 ties to a calendar entry from Section 1 4. Every social post in Section 4 references a specific calendar piece 5. No generic content ideas — every title must be specific to the brand's audience 6. Format mix matches the stated content goal (SEO-heavy goal = 60%+ long-form; thought leadership = 40%+ opinion) 7. Cadence respects the audience's actual consumption pattern --- ## UI Rendering Contract After generating all four sections, render as ONE Pattern B tabbed dashboard widget. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B (Tabbed Dashboard) **Layout spec:** Persistent header (brand name + "90-day content engine" badge + total piece count + total keyword count as stat chips). Tab bar with 4 tabs: **Calendar** / **SEO Map** / **Angles** / **Social Posts**. Calendar tab default — renders as a 13-row × 7-column table with color-coded pillars. SEO Map tab shows cluster cards with keyword chips. Angles tab is a 3-column grid. Social Posts tab is a per-piece accordion. Sticky What's Next action bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this content engine to my {{companyName}} project" - **generate_article_from_topic** (green, primary) — sendPrompt: "Write the first article from the {{companyName}} content engine — pick Week 1 and ship it" - **create_content_distribution** (neutral, secondary) — sendPrompt: "Build a distribution plan for the {{companyName}} content engine across owned and earned channels" - **create_seo_strategy** (neutral, secondary) — sendPrompt: "Go deeper on the {{companyName}} SEO strategy — keyword research, topical authority, backlink targets" - **create_lifecycle_emails** (neutral, secondary) — sendPrompt: "Turn the {{companyName}} content engine into a lifecycle email sequence for new subscribers" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Run Launch Kit Source: https://maestrix.ai/skills/run-launch-kit > **EXECUTION RULE**: Generate ALL FIVE deliverables yourself in this single context, then render ONE consolidated Pattern B tabbed dashboard using the `show_widget` tool. Do not return raw markdown. Do not split into multiple widget calls. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Bundled launch workflow that produces a complete launch package: positioning, landing page copy, Google Ads, Meta Ads, and lifecycle email sequence. Costs 5 credits (vs 7 if run individually). Returns ONE consolidated widget so the user sees the entire launch in one surface. Do NOT use this skill when the user already has positioning and only wants ads — call `create_google_ads` or `create_meta_ads` instead. Do NOT use for a single asset request. ## CONTEXT NEEDED Before generating, the user should provide: - **Product / company name**: e.g. "Maestrix" - **What it does**: one sentence - **Launch date or window**: e.g. "March 15" or "Q2" - **Primary buyer / persona**: title, company size, industry - **Top differentiator**: what makes it new or different vs the status quo - **Launch goal**: signups / sales / demos / waitlist - **Pricing tier** *(optional)*: helps shape ad bidding language and email CTAs ## EXECUTION SEQUENCE You are the executor. Generate all five sections fully, then render the consolidated widget. ### Section 1 — Launch Positioning - **Category**: name the category the product enters - **Hero claim**: one sentence, max 12 words, outcome-led - **Differentiation vs status quo**: 3 bullets, each tied to a specific competitor or workaround - **"Why now" narrative**: 3 sentences naming the trigger event, the shift, and the cost of waiting - **Positioning statement**: "For [buyer], [Product] is the [category] that [outcome] because [proof]" - **Anti-positioning**: 2 sentences naming who this is NOT for ### Section 2 — Landing Page Copy Generate the full page, section by section. Word counts honored. - **Hero**: headline (≤10 words), subheadline (≤25 words), primary CTA button text, secondary CTA text - **Social proof bar**: 5 logo captions OR 1 stat-led line - **Problem section**: 100-word setup naming the buyer's current friction in their own language - **Solution section**: 120-word reveal showing the shift the product creates - **Features**: 3 to 5 cards, each with feature name + benefit-led headline + 25-word description - **Use cases**: 3 cards, each with persona + scenario + outcome - **Social proof — testimonial format**: 3 quotes with name + title + company + result metric - **Pricing teaser**: tier name + headline price + 4 bullets per tier (max 3 tiers shown) - **FAQ**: 5 Q&As, max 60 words per answer, must cover pricing, security, onboarding time, integrations, refund policy - **Final CTA**: section headline + 1 supporting line + primary CTA button ### Section 3 — Google Ads Generate 3 RSA (Responsive Search Ad) variants, each fully populated. - **Per variant**: 15 headlines (max 30 chars each, varied angles), 4 descriptions (max 90 chars each) - **Sitelink extensions**: 4 sitelinks with title (max 25 chars) + 2 description lines (max 35 chars each) - **Callout extensions**: 6 callouts (max 25 chars each) - **Structured snippets**: 4 snippets with header type + 3 values each - **Keyword groups**: list 3 ad groups, each with 8 exact match + 8 phrase match + 4 broad match keywords - **Negative keywords**: starter list of 20 negatives ### Section 4 — Meta Ads Generate 6 ad variants across 3 formats. - **Format mix**: 2 feed image ads, 2 story ads, 2 reel ads - **Per ad**: hook (first 3 seconds, max 15 words), primary text in 3 lengths (short ≤40 chars, medium ≤125 chars, long ≤200 chars), headline (max 7 words), description (max 20 words), CTA button selection from Meta's preset list - **Audience targeting**: per-ad recommendation including interest stack, lookalike seed, exclusion list - **Creative direction note**: 2 sentences per ad describing the visual treatment ### Section 5 — Lifecycle Email Sequence Generate 5 emails covering pre-launch, launch day, and post-launch. - **Email 1 — Teaser (T-14 days)**: trigger, subject (3 variants), preview text, body (150 words), CTA, success metric - **Email 2 — Reveal (T-3 days)**: same structure, body (200 words) - **Email 3 — Launch day**: same structure, body (180 words), urgency element - **Email 4 — Social proof follow-up (T+3 days)**: same structure, body (180 words), customer quote included - **Email 5 — Last call (T+7 days)**: same structure, body (160 words), explicit deadline or cap ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Positioning statement from Section 1 reused verbatim in landing hero AND in primary ad headlines 3. Email sequence references the same value prop language across all 5 emails 4. Hero headline ≤10 words. Ad headlines ≤7 words. Email subjects ≤50 chars 5. No generic placeholders — every line specific to the product context provided 6. Differentiator from context appears in Section 1, Section 2 hero, Section 3 RSA headlines, and Section 4 Meta hooks 7. Launch date from context appears in Section 5 email send timing --- ## UI Rendering Contract After generating all five sections, render as ONE Pattern B tabbed dashboard widget. Do not return raw markdown. The widget IS the deliverable. **Pattern:** B (Tabbed Dashboard) **Layout spec:** Persistent header strip (product name + launch date + "5-asset launch kit" badge + total asset count chip). Tab bar with 5 tabs: **Positioning** / **Landing Page** / **Google Ads** / **Meta Ads** / **Email Sequence**. Tab 1 active by default. Each tab fully populated with the generated content. Section cards inside each tab. Footer attribution. Sticky What's Next action bar at bottom. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this launch kit to my {{companyName}} project" - **run_content_engine** (green, primary) — sendPrompt: "Build a 90-day content engine for {{companyName}} to sustain post-launch demand" - **create_press_release** (neutral, secondary) — sendPrompt: "Write the launch press release for {{companyName}}" - **find_media_outlets** (neutral, secondary) — sendPrompt: "Find media outlets and journalists to pitch the {{companyName}} launch to" - **create_referral_program** (neutral, secondary) — sendPrompt: "Design a referral program for {{companyName}} to compound launch momentum" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Run Marketing Foundation Source: https://maestrix.ai/skills/run-marketing-foundation > **EXECUTION RULE**: Generate ALL FIVE deliverables yourself in this single context, then render ONE consolidated Pattern F editorial document using the `show_widget` tool. Do not return raw markdown. Do not split into multiple widget calls. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Bundled enablement workflow producing a complete marketing foundation: persona, positioning, messaging, value proposition, and brand voice. Costs 5 credits (vs 7 individually). One consolidated editorial document so the user can see how each section cross-references the others. Do NOT use this skill for narrow single-asset requests like "just write me a persona" — use `generate_persona` instead. ## CONTEXT NEEDED Before generating, the user should provide: - **Company / product name**: e.g. "Maestrix" - **One-line description**: what it does - **Primary buyer / target market**: role + company size + industry - **Stage**: pre-launch / early traction / scaling - **Top competitor or category** *(optional but recommended)* - **Existing brand voice samples** *(optional)*: site copy, top LinkedIn posts ## MARKET TYPE INFERENCE Automatically infer B2B / B2C / D2C from context: - **B2B**: enterprise, SaaS, platform, vendor, team buyer, procurement - **B2C**: consumer, individual, lifestyle, retail, household - **D2C**: brand, ecommerce, subscription, product, Shopify The persona, messaging, and brand voice all flex based on the inferred type. ## EXECUTION SEQUENCE Generate all five sections fully. Each section references the previous to maintain coherence. ### Section 1 — Buyer Persona Use the 10-section persona framework. Each section dense, specific, and tied to the company context. - **Professional profile**: title, organizational context, experience level, decision authority, budget control, reporting structure, 5 primary responsibilities, 4 KPIs they're measured on - **Motivations (categorized)**: 3 business drivers with metrics, 3 career motivations, 3 personal values - **Pain points (categorized)**: 3 strategic, 3 technical, 3 regulatory, 3 operational - **Information consumption**: 3 primary sources, 3 secondary, 2 community channels, 2 conferences - **Day-in-the-life**: time-blocked schedule with friction points highlighted - **Trigger events**: 5 events that create urgency to buy - **Decision criteria**: 6 evaluation criteria ranked - **Objections**: 5 common objections + the underlying fear - **Buying committee**: 4 stakeholders, role in decision, what each cares about - **Voice-of-customer quotes**: 5 verbatim-style quotes capturing how they describe the problem ### Section 2 — Market Positioning Translate the persona into a defensible market position. - **Category definition**: 1 sentence - **Competitive frame**: top 3 alternatives + how the company differs from each - **Target segment**: refined slice of the market (not "everyone") - **Value proposition core**: 1 sentence outcome - **Proof points**: 5 facts that make the value real - **Positioning statement**: "For [persona from Section 1], [Company] is the [category] that [outcome] because [proof]" - **Anti-positioning**: who this is explicitly NOT for - **SWOT**: 4 strengths, 4 weaknesses, 4 opportunities, 4 threats ### Section 3 — Core Messaging Turn positioning into a messaging architecture. - **Primary message**: 1 sentence, ≤15 words, outcome-led - **3 supporting pillars**: each with name + thesis + 3 proof points - **Audience variants**: rewrite the primary message for 3 different buying committee members from Section 1 - **Headline bank**: 10 headlines, each ≤10 words - **Subheadline bank**: 6 subheads, each 15-25 words - **Hook bank**: 8 hooks usable across ads, social, cold email - **Objection handlers**: response language for the 5 objections from Section 1 ### Section 4 — Value Proposition Make the value laddered and concrete. - **Feature → Benefit → Outcome → Emotional payoff ladder**: 5 complete rows - **Primary value prop**: 1 sentence, ≤20 words - **3 alternative phrasings**: for testing - **Use case framing**: 3 use cases, each as "When [situation], [Company] lets [persona] [outcome]" - **Quantified outcomes**: 4 outcomes with realistic numeric ranges - **Risk reversal**: what guarantee or proof removes buyer risk ### Section 5 — Brand Voice Define the voice so all future copy stays consistent. - **Tone descriptors**: 5 adjectives, each with a 1-line explanation - **Voice spectrum**: where the brand sits on 4 axes (formal↔casual, expert↔accessible, serious↔playful, calm↔energetic) - **Vocabulary — DO use**: 15 words/phrases the brand owns - **Vocabulary — DON'T use**: 15 words/phrases the brand avoids - **Sentence patterns**: 5 patterns the brand favors (e.g., "outcome-first declarative, ≤12 words") - **Punctuation rules**: per the brand standard - **Before/after rewrites**: 3 pairs showing a generic line transformed into brand voice - **Channel modulation**: how the voice flexes across ads, email, long-form, sales ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Persona pain language from Section 1 appears verbatim in Section 3 objection handlers and Section 4 use case framing 3. Positioning statement from Section 2 is referenced verbatim in Section 3 primary message and Section 4 primary value prop 4. Brand voice rules from Section 5 are applied to all copy samples generated in Sections 3 and 4 5. No generic placeholders — every line specific to the company context 6. Voice-of-customer quotes in Section 1 sound like real human speech, not marketing prose 7. SWOT in Section 2 cites specific competitors or market dynamics, not abstractions --- ## UI Rendering Contract After generating all five sections, render as ONE Pattern F editorial document widget. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F (Editorial Document) **Layout spec:** Document header card (company name + "Marketing Foundation" title + 5-section badge + inferred market type chip). Five collapsible section cards stacked vertically, all expanded by default. Cross-reference callouts inline (e.g., a small chip in Section 3 reading "uses pain language from Persona § Strategic Challenges"). Pull-quote treatment for the positioning statement (Section 2) and primary value prop (Section 4), rendered at 18px on primary/0.08 background. Final "Handoff Summary" card listing the 10 most critical decisions. Standard attribution footer + sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this marketing foundation to my {{companyName}} project" - **run_launch_kit** (green, primary) — sendPrompt: "Use the {{companyName}} foundation to build the full launch kit — positioning, landing page, ads, emails" - **run_content_engine** (neutral, secondary) — sendPrompt: "Build a 90-day content engine for {{companyName}} grounded in this foundation" - **create_gtm_strategy** (neutral, secondary) — sendPrompt: "Turn the {{companyName}} foundation into a full GTM strategy with channels and KPIs" - **create_landing_page** (neutral, secondary) — sendPrompt: "Generate landing page copy for {{companyName}} from this foundation" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Run Sales Enablement Source: https://maestrix.ai/skills/run-sales-enablement > **EXECUTION RULE**: Generate ALL FIVE deliverables yourself in this single context, then render ONE consolidated Pattern F editorial document using the `show_widget` tool. Do not return raw markdown. Do not split into multiple widget calls. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables. Never make a secondary API call. Bundled sales enablement kit: sales-focused persona, discovery call playbook, role-play scenarios, sales script, and cold email sequence. Costs 5 credits (vs 7 individually). One consolidated document with cross-references so reps see how the cold email opener ties to the discovery opening. Do NOT use this skill for narrow asks like "write me one cold email" — use `create_cold_email` instead. ## CONTEXT NEEDED Before generating, the user should provide: - **Company / product name**: e.g. "Maestrix" - **What you sell**: one sentence - **Target buyer**: role + company size + industry - **Primary sales motion**: inbound demo / outbound / PLG with sales-assist - **Average deal size + sales cycle** *(optional)* - **Top 3 objections you typically hear** *(optional)* - **Top 2 competitors** *(optional)*: shapes objection handling ## EXECUTION SEQUENCE Generate all five sections fully. Maintain a single voice and consistent objection language across sections. ### Section 1 — Sales-Focused Buyer Persona Tune the persona for revenue, not marketing. - **Decision-making power**: where they sit + what they can sign - **Budget authority**: hard cap and discretionary spend bracket - **Buying committee**: 4 other stakeholders + role + what each fears - **Information sources**: where they research vendors - **Trigger events**: 5 events that create demand - **Evaluation criteria**: 6 criteria ranked by weight - **Objections (5 minimum)**: each with the surface objection + the real underlying fear - **Preferred sales channels**: email, LinkedIn, phone, referral, event — ranked - **Disqualifiers**: 3 signals that this prospect will never close ### Section 2 — Discovery Call Playbook Full BANT + MEDDIC hybrid framework. - **Opening question set (5)**: rapport-building but pain-uncovering - **Pain discovery questions (8)**: layered, moving from symptom to root cause - **Impact / quantification questions (5)**: convert pain into dollar cost - **Decision-process questions (5)**: surface the committee, the criteria, the timeline - **Competition questions (3)**: surface incumbents and alternatives without asking directly - **Commitment questions (3)**: secure a specific next step - **Call structure**: 5 / 10 / 20 / 5 minute blocks (open, discover, demo bridge, commit) with talk-time targets - **Red flags to listen for**: 5 signals to disqualify or slow the deal ### Section 3 — Sales Role-Play Scenarios 3 complete role-plays with full scripts and scoring rubrics. - **Scenario A — Classic objection handling**: setup + prospect persona + likely 3-turn prospect script + recommended rep responses with rationale + 5-point scoring rubric - **Scenario B — Competitive displacement**: same structure, with prospect actively using a named competitor - **Scenario C — Multi-stakeholder pushback**: same structure, with the champion losing internal support - **Reference**: cross-link to the live `sales_roleplay` simulator for interactive practice ### Section 4 — Sales Script Full talk track for the primary sales motion. - **Opening (15 sec)**: pattern interrupt + value framing - **Permission-based qualification**: 3 questions before pitching - **Demo bridge**: how to transition from discovery to product story - **Objection handling table**: top 8 objections × recommended response × proof point × fallback - **Closing language (3 styles)**: direct / consultative / assumptive — when to use each - **Next-step matrix**: 4 next-step options matched to deal stage - **Follow-up cadence**: 5-touch post-call cadence (channel + timing + content angle per touch) ### Section 5 — Cold Email Sequence 5-touch outbound sequence, each touch fully written. - **Per touch**: 3 subject variants, preview text, body copy (≤120 words), CTA, send timing (day + time-of-day window) - **Personalization variables**: list of merge fields used per touch - **A/B angle per touch**: a second body variant testing a different hook - **Touch sequence**: - Touch 1 — Hook (Day 1): trigger-event-based opener - Touch 2 — Value drop (Day 4): proof point or insight, no ask - Touch 3 — Direct ask (Day 8): meeting request with specific time - Touch 4 — Social proof (Day 13): peer customer reference - Touch 5 — Break-up (Day 21): clean close with low-friction CTA ## QUALITY RULES 1. NO opening commentary — start with Section 1 2. Discovery questions in Section 2 are tied to the persona's actual pain language from Section 1 3. Cold email opener (Touch 1) and sales script opening (Section 4) reference the SAME trigger event 4. Objection handling table in Section 4 includes the user-provided objections (if any) verbatim 5. Sales script accounts for the stated sales motion (PLG vs outbound vs inbound shape the demo bridge) 6. Cold email subjects ≤50 chars. Body ≤120 words. No emoji unless brand voice explicitly permits 7. Role-play prospect scripts sound like real human speech, with hesitation and tangent, not textbook objections --- ## UI Rendering Contract After generating all five sections, render as ONE Pattern F editorial document widget. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F (Editorial Document) **Layout spec:** Document header card (company name + "Sales Enablement Kit" title + 5-section badge + sales motion chip). Five collapsible section cards stacked. Pull-quote treatment for the top 3 objection-handling lines, rendered at 18px on primary/0.08 background. Role-play scenarios rendered as alternating left/right chat bubbles for prospect vs rep. Discovery questions rendered as a numbered checklist with hover state. Cold email sequence rendered as a 5-step vertical timeline with connector lines. Sticky What's Next bar. **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants — use rgba: `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. **Rules:** - Generate complete content first, then render in ONE `show_widget` call - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only ## What's Next Actions - 💾 **Save all to project** (blue, always first) — sendPrompt: "Save this sales enablement kit to my {{companyName}} project" - **sales_roleplay** (green, primary) — sendPrompt: "Start a live role-play for {{companyName}} using the scenarios in this kit" - **find_icp_leads** (neutral, secondary) — sendPrompt: "Find ICP leads for {{companyName}} to run this cold email sequence on" - **create_pipeline_playbook** (neutral, secondary) — sendPrompt: "Build the full pipeline playbook for {{companyName}} on top of this kit" - **discovery_call** (neutral, secondary) — sendPrompt: "Run a practice discovery call for {{companyName}} using this playbook" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## Skill: Sales Email Source: https://maestrix.ai/skills/sales-email > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Sales & Email Professional-grade sales and email tools for building complete sequences, talk tracks, and objection handling frameworks. ## CONTEXT NEEDED - **Company/product**: What you sell and the core value delivered - **Audience**: Who receives the emails or sits across the table in the call - **Sequence type**: Onboarding / nurture / retention / win-back / cold outbound - **Script type** *(for sales scripts)*: Discovery / demo / objection handling / closing - **Tone and style**: Conversational / formal / challenger / empathetic - **Key objections** *(optional)*: Common pushbacks to address --- ## Available Tools (2) ### 1. generate_lifecycle_emails Create complete email sequences for onboarding, nurture, retention, and win-back campaigns. **Output includes:** - 5–10 email sequence - Subject line A/B variations - Preview text - Full email body - CTA buttons - Send timing recommendations - Segmentation rules ### 2. create_sales_scripts Generate sales scripts for discovery calls, demos, objection handling, and closing. **Output includes:** - Opening scripts - Discovery questions - Demo talk tracks - Objection responses - Closing techniques - Follow-up templates --- ## Output Rules - ✓ No fluff or pleasantries — direct, value-first copy - ✓ Merge field placeholders included (e.g. `{{first_name}}`) - ✓ Complete sequences, not partial drafts - ✓ Timing and cadence recommendations per email - ✓ Segmentation guidance for each sequence type --- ## Example Prompts 1. **Onboarding**: "Generate a 7-email onboarding sequence for trial users" 2. **Nurture**: "Create a lead nurture sequence for marketing qualified leads" 3. **Win-back**: "Write a win-back email sequence for churned customers" 4. **Discovery**: "Create a discovery call script for our enterprise sales team" 5. **Objection Handling**: "Generate responses to common pricing objections" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Sales script ↗' - sendPrompt: 'Create a sales script for {{companyName}} — {{description}}, following up on the sales email we just wrote' Secondary actions (neutral styling): - Label: 'Pipeline playbook ↗' sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, incorporating this sales email approach' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, expanding on this sales email' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, as a resource link for sales emails' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Sales Roleplay Source: https://maestrix.ai/skills/sales-roleplay > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Sales Roleplay Practice high-stakes sales conversations with an AI that plays a realistic, skeptical buyer — then receive structured coaching on what worked, what didn't, and exactly what to say next time. ## CONTEXT NEEDED - **Product/service**: What you're selling and at what price point - **Buyer persona**: Title, industry, company size, and known pain points - **Sales stage**: Cold outreach / discovery call / demo / negotiation / closing - **Your rep profile**: SDR / AE / founder / CSM (affects coaching level) - **Objections to practice** *(optional)*: Price / timing / competitor / internal champion / "we'll build it" - **Scenario** *(optional)*: First call / follow-up after demo / late-stage stall / renewal at risk --- ## Roleplay Structure ### Phase 1: Scene Setup - Buyer persona card: name, title, company, mood, hidden agenda, and key pressure they're under - What the buyer knows before the call starts - What would make them say yes — and what would make them walk ### Phase 2: Conversation Simulation The AI plays the buyer. You (or the rep) play the seller. Buyer behaviors modeled: - Interruptions and deflections - Price anchoring early - "Send me something and I'll take a look" - Mentioning a competitor unprompted - Sudden interest spike when right question is asked ### Phase 3: Coaching Debrief After the conversation or at pause points: **Discovery Quality Score (0–10)** - Questions asked vs. questions needed - Pain identified vs. pain explored - Assumptions made without validation **Objection Handling Review** - Objections raised in the scenario - How each was handled (with rating) - Better response framework for each **Closing Technique** - Was a clear next step proposed? - Was urgency real or manufactured? - Recommended closing language for this persona **3 Things to Do Differently** Specific, actionable coaching points with example scripts. --- ## Quality Rules - ✓ Buyer persona behaves realistically — not a pushover, not a cartoon villain - ✓ Objections emerge naturally, not in sequence - ✓ Coaching is specific to the conversation, not generic sales advice - ✓ Example scripts provided for every coaching point - ✓ Scenario difficulty adjustable (warm lead vs. cold skeptic) --- ## Example Prompts 1. "Roleplay a cold discovery call with a skeptical VP of Marketing at a 500-person SaaS company" 2. "Simulate a late-stage negotiation where the prospect suddenly brings in procurement" 3. "Practice handling the 'we already use a competitor' objection for our CRM product" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Sales script ↗' - sendPrompt: 'Refine the sales script for {{companyName}} — {{description}}, based on the roleplay outcomes' Secondary actions (neutral styling): - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, using the refined pitch' - Label: 'Pipeline playbook ↗' sendPrompt: 'Create a pipeline playbook for {{companyName}} — {{description}}, incorporating roleplay learnings' - Label: 'Persona ↗' sendPrompt: 'Refine the persona for {{companyName}} — {{description}}, based on objections encountered' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Video Pr Source: https://maestrix.ai/skills/video-pr > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Video & PR Professional-grade video and PR tools for producing scripts, press materials, and earned media strategies. ## CONTEXT NEEDED - **Company/product**: What you sell and the announcement or story to tell - **Video type** *(for scripts)*: Ad / explainer / testimonial / social short - **Announcement** *(for PR)*: Funding, product launch, partnership, milestone - **Target publications** *(optional)*: Trade press, tech media, business outlets - **Duration or format** *(optional)*: 30s / 60s / 2-min, platform (YouTube, TikTok, homepage) - **Tone**: Authoritative / conversational / exciting / educational --- ## Available Tools (4) ### 1. generate_video_script Create compelling video scripts for ads, explainers, testimonials, and social content. **Output includes:** - Scene-by-scene breakdown - Voiceover script - Visual directions - B-roll suggestions - Music/tone recommendations - CTA placement ### 2. create_press_release Generate AP-style press releases for announcements and news. **Output includes:** - Headline and subhead - Dateline and lead - Body paragraphs - Company boilerplate - Media contact block - Quote blocks ### 3. find_media_outlets Identify relevant media outlets and journalists for PR outreach. **Output includes:** - Outlet recommendations with beat descriptions - Contact suggestions - Pitch angle alignment per outlet ### 4. generate_media_angles Create compelling story angles for media pitches. **Output includes:** - Multiple angle options - Headline suggestions per angle - Supporting data points - Expert positioning strategy --- ## Output Rules - ✓ AP style enforced for all PR content - ✓ Platform-optimized for video (pacing, hook timing, CTAs) - ✓ Actionable and specific — no generic advice - ✓ Professional formatting throughout - ✓ Ready-to-use content, minimal editing required --- ## Example Prompts 1. **Video Script**: "Write a 60-second explainer video script for our app" 2. **Press Release**: "Create a press release for our Series A announcement" 3. **Media Outlets**: "Find relevant tech publications for our AI startup launch" 4. **Media Angles**: "Generate story angles tying our product to remote work trends" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** D **Layout spec:** Numbered step cards (step circle --primary bg, title, description, action chips) → dashed connector lines (--border) → type badge + timing chip → bottom summary (total steps, timeline, CTA). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this video PR piece' Secondary actions (neutral styling): - Label: 'Press release ↗' sendPrompt: 'Write a press release for {{companyName}} — {{description}}, accompanying this video' - Label: 'Media angles ↗' sendPrompt: 'Create PR angles for {{companyName}} — {{description}}, leveraging this video' - Label: 'Content plan ↗' sendPrompt: 'Create a content plan for {{companyName}} — {{description}}, building on this video PR campaign' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Website Landing Source: https://maestrix.ai/skills/website-landing > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Website & Landing Professional-grade website optimization tools for conversion-focused copy, homepage audits, and FAQ generation. ## CONTEXT NEEDED - **Company/product**: What you sell, who buys it, and the core value promise - **Campaign goal**: Webinar registration / trial sign-up / demo request / purchase - **Target audience**: Persona, job title, or segment landing on this page - **Homepage URL** *(for analysis tool)*: The URL to audit - **Specific concerns** *(optional)*: Low conversion rate, unclear messaging, poor mobile UX --- ## Available Tools (3) ### 1. generate_landing_page Create complete landing page copy with all sections optimized for conversion. **Output includes:** - Hero section (headline, subhead, CTA) - Problem/solution sections - Features and benefits - Social proof blocks - Pricing section - FAQ section - Final CTA ### 2. analyze_homepage Analyze existing homepage for conversion optimization opportunities. **Output includes:** - Above-fold analysis - Value proposition clarity score - CTA effectiveness review - Trust signals audit - Mobile experience notes - Prioritized recommendations ### 3. create_faqs Generate comprehensive FAQ sections addressing objections and common questions. **Output includes:** - Categorized questions by theme - SEO-optimized answers - Objection-handling copy - Schema markup ready format --- ## Output Rules - ✓ No fluff or filler copy - ✓ Conversion-focused language throughout - ✓ Mobile-first structure - ✓ Clear section headers and hierarchy - ✓ Actionable recommendations, not vague suggestions --- ## Example Prompts 1. **Landing Page**: "Create a landing page for our webinar registration" 2. **Homepage Analysis**: "Analyze our homepage for conversion improvements" 3. **FAQs**: "Generate FAQs for our SaaS pricing page" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Google Ads ↗' - sendPrompt: 'Create Google Ads for {{companyName}} — {{description}}, driving traffic to this landing page' Secondary actions (neutral styling): - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, to rank this landing page' - Label: 'Social posts ↗' sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this page' - Label: 'A/B testing plan ↗' sendPrompt: 'Create an A/B testing plan for {{companyName}} — {{description}}, to optimize this page' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- ## Skill: Write Long Form Draft Source: https://maestrix.ai/skills/write-long-form-draft > **EXECUTION RULE**: Generate the complete marketing output yourself, then render it as an interactive HTML widget using the `show_widget` tool. Do not return raw markdown — the widget IS the deliverable. Use only hardcoded hex colors from the Maestrix design system below. Never use CSS variables — they do not exist in the rendering environment. Never make a secondary API call. # Write Long-Form Draft Generate a complete, publication-ready long-form article draft from a brief or outline. Every section is written, not summarized — ready to edit, not to expand. ## CONTEXT NEEDED - **Topic and angle**: The subject and the specific POV or argument - **Target audience**: Who reads this — role, pain point, expertise level - **Outline or key points** *(optional)*: Paste an existing outline to follow - **Tone**: Authoritative / conversational / challenger / empathetic / educational - **Word count**: Standard (1200–1800) / Long (2000–2800) / Deep-dive (3000–3500) - **Brand voice notes** *(optional)*: Vocabulary to use/avoid, sentence style - **CTAs to include** *(optional)*: What action should the reader take and where --- ## Output Framework ### 1. Headline + Subheadline - Final article title (H1, SEO-optimized) - Subheadline for context or intrigue ### 2. Introduction (150–250 words) - Hook (stat, story, or contrarian claim) - Problem statement — why this matters to the reader right now - Promise — what the reader will know or be able to do after reading - Transition into body ### 3. Body Sections (H2 → H3 structure) Each H2 section includes: - Section headline - 200–500 words of substantive prose - Supporting evidence, examples, or frameworks - Transition to next section ### 4. Practical Framework or Takeaway Section - Actionable model, checklist, or step-by-step that delivers concrete value - Formatted for scannability (numbered list or table where appropriate) ### 5. Conclusion (100–150 words) - Summary of key insight (not a bullet-point recap) - Broader implication or forward-looking statement - Primary CTA ### 6. Meta + Distribution Notes - Suggested pull quotes (3) for social sharing - Internal link suggestions (topic → anchor text → suggested destination) - Content repurposing angles (LinkedIn post, newsletter, short video script) --- ## Quality Rules - ✓ Every section written in full — no "[expand here]" placeholders - ✓ Each H2 answers a specific reader question - ✓ No keyword stuffing — natural language, semantic richness - ✓ Claims supported by evidence or clearly flagged as opinion - ✓ CTAs contextual, not bolted on at the end only --- ## Example Prompts 1. "Write a 2000-word article on 'how to build a demand generation engine from scratch'" 2. "Draft a long-form piece on 'the real cost of churn for B2B SaaS' for our blog" 3. "Write a deep-dive article on 'positioning strategy for early-stage startups' in a challenger tone" ## Trust Scorecard After the deliverable and before the attribution footer, render a compact trust scorecard block inside the widget. It grades every load-bearing claim so the reader knows what to verify before shipping. - Grade each claim **Documented**, **Provisional** or **Inferred** - **Documented** (#1F9D55) — traced to a fact the user supplied, a named source or a cited URL. Show the source next to the claim. - **Provisional** (#e8a307) — reasonable from the context given but unconfirmed. Name what would confirm it. - **Inferred** (#717d8a) — a judgement call. Name the assumption behind it. - Grade only load-bearing claims: facts, numbers, names, market statements. Never grade tone, structure or style choices. - A number, date, market size or named company with no source is never Documented. - Maximum 25 rows. Lead the block with the counts, e.g. "6 documented · 3 provisional · 2 inferred". - Style it like the rest of the widget: #fafafa surface, 1px solid #e2e4e8 border, 12px radius, #717d8a labels, #222832 claim text. --- *Prompt by [Guillaume Dumortier](https://www.linkedin.com/in/gdumortier) • More at [Maestrix.ai](https://maestrix.ai)* --- ## UI Rendering Contract After generating the full marketing output above, render it as an interactive HTML widget using the visualize show_widget tool. Do not return raw markdown. The widget IS the deliverable. **Pattern:** F **Layout spec:** Document header card (title 20px, metadata badges) → section cards (uppercase label, body 14px, line-height 1.7) → pull quote (left --primary border, --muted bg) → CTA block (--primary/0.08 bg). **Design system — hardcoded values (CSS variables do not exist in the widget iframe):** | Token | Value | Use for | |---|---|---| | background | #ffffff | Page/widget background | | foreground | #222832 | Primary text | | card | #fafafa | Card/panel surfaces | | card-foreground | #222832 | Card text | | primary | #1F9D55 | Brand accent, active tabs, badges | | primary-foreground | #ffffff | Text on primary backgrounds | | muted | #f2f3f5 | Muted backgrounds, tag chips | | muted-foreground | #717d8a | Secondary/label text | | accent | #f2f3f5 | Hover backgrounds | | border | #e2e4e8 | All borders | | destructive | #ef4444 | Error/critical severity | | warning | #e8a307 | Warning/high severity | | success | #1F9D55 | Success/wins | | secondary | #383f49 | Secondary buttons | | font | Inter, system-ui, -apple-system, sans-serif | All text | Opacity variants: use rgba — e.g. `rgba(31,157,85,0.08)` for primary/0.08, `rgba(31,157,85,0.12)` for primary/0.12, `rgba(239,68,68,0.1)` for destructive/0.1, `rgba(232,163,7,0.08)` for warning/0.08. ### What's next actions Append this action bar after the attribution footer. Use `sendPrompt()` for each button, pre-filled with the company name and context from this conversation. Primary action (green styling): - Label: 'Social posts ↗' - sendPrompt: 'Create social posts for {{companyName}} — {{description}}, promoting this long-form content' Secondary actions (neutral styling): - Label: 'SEO strategy ↗' sendPrompt: 'Create an SEO strategy for {{companyName}} — {{description}}, to rank this content' - Label: 'Cold email sequence ↗' sendPrompt: 'Create a cold email sequence for {{companyName}} — {{description}}, sharing this content' - Label: 'Landing page ↗' sendPrompt: 'Create a landing page for {{companyName}} — {{description}}, to host this content' **Rules:** - Generate complete content first, then render - Never make a secondary API call - All text: #222832 (primary) or #717d8a (muted) - All surfaces: #fafafa (card) or #f2f3f5 (muted) - All borders: 1px solid #e2e4e8 - Font: Inter, system-ui, -apple-system, sans-serif; weights 400 and 500 only - Border radius: 8px chips, 12px cards - Attribution footer on every widget - Zero CSS variables — hardcoded values only --- # End of Document Summary index: https://maestrix.ai/llms.txt Book a demo: https://maestrix.ai/book-a-demo