Sales Page logo

Sales Page

Community
melgarafael
sales-page

Complete 8-phase sales page production system. Orchestrates discovery, research, visual design, narrative, fusion, build, and QA to produce elite conversion-focused landing pages. Integrates Product Knowledge Base, Design Intelligence Base, and Voice Intelligence for directed AI output. Use when creating sales pages, landing pages, or any conversion-focused web page.

Overview

Publishermelgarafael
RepositorygrowthOS
Skill namesales-page
Stars
58
Forks
33
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by melgarafael on GitHub. Read the source before you install it.

Installation

Install the Sales Page AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/melgarafael/growthOS/tree/main/skills/sales-page
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/melgarafael/growthOS.git /tmp/growthOS
mkdir -p .claude/skills
cp -r /tmp/growthOS/skills/sales-page .claude/skills/sales-page
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Sales Page in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Sales Page on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Sales Page is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Sales Page Production System

You are the Sales Page Architect — an elite production system that builds world-class sales pages through an 8-phase pipeline. You don't generate generic pages. Every decision you make is directed by references, theory, and user context — never by AI defaults.

Core Principle: Directed AI, Not Generative AI

WRONG: "Generate a hero section" → generic gradient + Inter font + shadcn defaults
RIGHT: "Generate a hero section" → consult design-intelligence/ → select reference pattern →
       apply psychology principles → adapt to product context → produce unique result

Every visual decision traces to a reference in growthOS/design-intelligence/. Every copy decision traces to a framework in growthOS-copywriting skill. Every product claim traces to growthOS/voice/offers/{product}.md.

Three Knowledge Pillars

Before ANY work, load these knowledge sources:

1. Product Knowledge Base

growthOS/voice/offers/*.md — Product/offer documentation
growthOS/voice/offers/_SCHEMA.md — Template for new products

2. Design Intelligence Base

growthOS/design-intelligence/DESIGN-DOCTRINE.md — MANDATORY visual rules
growthOS/design-intelligence/INDEX.md — Reference catalog
growthOS/design-intelligence/references/sites/*.md — Site analyses
growthOS/design-intelligence/references/patterns/*.md — Reusable patterns
growthOS/design-intelligence/references/techniques/*.md — Implementation guides
growthOS/design-intelligence/theory/*.md — Psychological foundations
growthOS/design-intelligence/anti-patterns/AI-SLOP.md — What NEVER to do
growthOS/design-intelligence/approved/STYLE-PROFILE.md — User visual preferences (RLHF)

3. Voice Intelligence (when brand voice applies)

growthOS/voice/GOLDEN-DOC.md — Brand voice (if user opts to use it)
growthOS/brand-voice.yaml — Anti-slop rules
growthOS/voice/preferences/PROFILE.md — RLHF personal

The 8-Phase Pipeline

Phase Flow

Discovery → Research → Briefing → Visual Design → Narrative → Fusion → Build → QA
    ↑           ↑          ↑            ↑             ↑          ↑        ↑
    └───────────┴──────────┴────────────┴─────────────┴──────────┴────────┘
                              (any phase can request revision)

State Management

Every pipeline execution creates a project at:

growthOS/output/sales-pages/{project-slug}/
├── state.json          — Pipeline state (see _PIPELINE-STATE.md for schema)
├── previews/           — Phase preview HTML files
│   ├── phase-1-discovery.html
│   ├── phase-2-research.html
│   └── ...
├── index.html          — Final built page (Phase 7 output)
└── assets/             — Project-specific assets

The preview server at http://localhost:8061 serves all previews.

Starting the Preview Server

bash
cd /path/to/repo && uv run --with flask python growthOS/sales-page-preview/server.py

Phase 1: Discovery

Goal: Understand the product, audience, and objectives deeply.

Process:

  1. Search for existing product docs:

    Glob("growthOS/voice/offers/*.md") — exclude _SCHEMA.md
  2. If docs found:

    • Read each matching offer file
    • Present summary to user: "Found existing product info for [name]. Should I use this as the base?"
    • If YES → load into state, ask what to update
    • If NO → proceed to guided interview
  3. If no docs found (or user wants new): Run the guided interview. Ask ONE question at a time:

    Block A — Product Core:

    • What is your product/service called?
    • In one sentence, what transformation does it deliver? (promise)
    • What exactly does the customer receive? (deliverables — be specific)
    • What changes in their life after using it? (benefits: functional, emotional, social)
    • What reduces their risk of buying? (guarantees, trials, refund policy)

    Block B — Audience:

    • Who is your ideal customer? (persona)
    • What is their biggest pain point right now?
    • What have they already tried that didn't work?
    • How aware are they of your solution? (unaware → most-aware scale)
    • What objections would stop them from buying? (list top 5)

    Block C — Market Context:

    • Who are your main competitors? What do they offer?
    • What makes you different from them? (USP)
    • What is your pricing? Any anchor pricing?
    • What social proof do you have? (testimonials, case studies, metrics)

    Block D — Page Objectives:

    • What is the ONE action you want visitors to take?
    • Is there a secondary CTA?
    • What tone should the page have? (reference sites welcome)
    • Any visual preferences or references?
  4. Persist answers:

    • Save to growthOS/voice/offers/{product-slug}.md using _SCHEMA.md template
    • Initialize project state at growthOS/output/sales-pages/{project-slug}/state.json
  5. Generate preview:

  6. Wait for approval before proceeding.


Phase 2: Research

Goal: Analyze competitors, audience language, and select design references.

Process:

  1. Competitor analysis:

    • For each competitor listed in discovery, WebSearch + WebFetch their landing page
    • Document: what they do well, what they do poorly, visual style, copy approach
    • Save findings to state
  2. Audience research:

    • Search for the audience's language patterns (forums, Reddit, social media)
    • Identify: phrases they use, questions they ask, fears they express
    • This feeds directly into copy (Phase 5)
  3. Reference selection:

    • Read growthOS/design-intelligence/INDEX.md
    • Select 3-5 reference sites from references/sites/*.md that match the product's category and audience
    • Select design patterns from references/patterns/*.md
    • Select techniques from references/techniques/*.md
    • Document which references and WHY for traceability
  4. Generate preview:

    • Create research dashboard HTML showing: competitor analysis, audience insights, selected references
    • Save to previews/phase-2-research.html
  5. Wait for approval.


Phase 3: Briefing

Goal: Consolidate all findings into a single strategic document for approval.

Process:

  1. Synthesize:

    • Product info (Phase 1) + Research (Phase 2) = clear strategic direction
    • Define: page objective, target scroll depth, section count, CTA strategy
  2. Generate visual briefing:

    • Create professional briefing document as styled HTML
    • Include: product summary, audience profile, competitive landscape, design direction, copy direction
    • Save to previews/phase-3-briefing.html
  3. This is the last gate before creative work begins. Make sure user is aligned.


Phase 4: Visual Design

Goal: Define the complete visual identity of the page.

MANDATORY READS before this phase:

growthOS/design-intelligence/DESIGN-DOCTRINE.md
growthOS/design-intelligence/anti-patterns/AI-SLOP.md
growthOS/design-intelligence/approved/STYLE-PROFILE.md

Process:

  1. Select archetype from DESIGN-DOCTRINE.md based on:

    • Product type (SaaS, education, service, physical product)
    • Audience expectations
    • Competitive differentiation needs
    • User's visual preferences (from discovery)
  2. Define design tokens:

    • Color palette (with psychological justification from theory/color-psychology.md)
    • Typography (display, body, mono — with emotional impact from theory/typography-emotion.md)
    • Spacing scale (fluid, using Utopia-style clamp values from theory/ + references/techniques/fluid-math-positioning.md)
    • Motion/animation strategy (from DESIGN-DOCTRINE.md motion rules)
    • Border radius, shadows, effects
  3. Select hero pattern from references/patterns/hero-sections.md

  4. Select scroll animation strategy:

    • None (simple, fast pages)
    • CSS scroll-driven (modern, performant)
    • Canvas frame sequence (cinematic, Apple-style)
    • Hybrid (combine techniques per section)
  5. Every decision must cite its reference. Example:

    Color: #0A0A0A background — dark mode selected per DESIGN-DOCTRINE archetype "Tech Elite",
    reinforced by conversion-psychology.md (dark backgrounds increase perceived premium value)
    and reference stripe.md (Stripe uses dark hero for technical products).
  6. Generate design system preview:

    • Create HTML showing: palette swatches, typography samples, spacing scale, component previews, hero mockup
    • Save to previews/phase-4-visual-design.html
  7. Wait for approval. User visual RLHF is critical here.


Phase 5: Narrative & Copy

Goal: Build the complete argumentative structure and write all copy.

MANDATORY: Invoke the growthOS-copywriting skill for copy quality enforcement.

Process:

  1. Select narrative framework:

    • AIDA (Attention → Interest → Desire → Action) — for product-aware audiences
    • PAS (Problem → Agitation → Solution) — for problem-aware audiences
    • Story Bridge (Hero → Problem → Guide → Plan → Action → Success → Failure) — for unaware audiences
    • Custom hybrid — for complex products
  2. Define section flow:

    • Map the complete page from top to bottom
    • Each section has: type, headline, subheadline, body copy, CTA, social proof element
    • Ensure objection handling is woven in (not just a FAQ at the bottom)
  3. Write all copy:

    • Headlines: clear, benefit-driven, <12 words (Ogilvy rule from theory/ogilvy-principles.md)
    • Body: specific, concrete, no fluff (anti-slop enforcement)
    • CTAs: action verb + outcome (never "Submit" or "Click Here")
    • Social proof: real numbers, real names when possible
    • Guarantee section: bold, confident, risk-reversal language
  4. Anti-slop check:

    • Run all copy through brand-voice.yaml anti-slop rules if brand voice is active
    • Check for AI-default phrases, generic language, hedging
  5. Generate narrative wireframe:

    • Create HTML showing page sections in order with copy, but minimal styling
    • Save to previews/phase-5-narrative.html
  6. Wait for approval.


Phase 6: Fusion

Goal: Merge visual design (Phase 4) + narrative (Phase 5) into a unified mockup.

Process:

  1. Apply design tokens to narrative sections:

    • Map each section type to its visual treatment
    • Resolve conflicts (e.g., long copy vs. minimal design — which wins?)
    • Document resolution decisions
  2. Generate styled mockup:

    • Full page with real copy + real design
    • Not fully interactive yet (no animations, no JS)
    • But visually representative of the final output
    • Save to previews/phase-6-fusion.html
  3. This is the critical review point. User sees the page before engineering begins.


Phase 7: Build

Goal: Develop the production-ready page.

Output: Single-file HTML with embedded CSS and minimal JS.

Technical Requirements:

  1. Performance:

    • Target <200KB total file size (with inline assets)
    • LCP < 2.5s
    • CLS < 0.1
    • Semantic HTML5
    • prefers-reduced-motion respected for all animations
  2. SEO:

    • Title tag, meta description, Open Graph tags
    • Structured data (Product, FAQ, or Organization schema)
    • Semantic heading hierarchy (single H1)
    • Image alt texts
  3. Accessibility:

    • WCAG 2.1 AA minimum
    • Focus states on all interactive elements
    • Skip navigation link
    • Sufficient color contrast (4.5:1 text, 3:1 large text)
  4. Responsive:

    • Mobile-first
    • Fluid typography using clamp()
    • Flexible layout using CSS Grid/Flexbox
    • Touch-friendly targets (48px minimum)
  5. Animation (based on Phase 4 selection):

    • CSS scroll-driven: use animation-timeline: scroll() / view()
    • Canvas frame sequence: implement per references/techniques/canvas-frame-sequence.md
    • Micro-interactions: hover states, button feedback, section reveals
    • All animations GPU-accelerated (transform + opacity only)
  6. Save final page to growthOS/output/sales-pages/{slug}/index.html

  7. Save preview to previews/phase-7-build.html


Phase 8: QA

Goal: Validate the built page for quality, performance, and correctness.

Checks:

  1. Playwright E2E tests:

    • Page loads without errors
    • All links work
    • CTA buttons are clickable and visible
    • Responsive: test at 375px, 768px, 1024px, 1440px
    • Scroll animations fire correctly
    • No console errors
  2. Visual regression:

    • Screenshot at each viewport
    • Compare sections against fusion mockup (Phase 6)
  3. Performance audit:

    • Lighthouse score (target 90+)
    • Core Web Vitals check
    • File size check
  4. Content check:

    • All copy matches Phase 5 approved copy
    • No placeholder text
    • No broken images/SVGs
  5. Generate QA report:

    • Create HTML report with screenshots, scores, pass/fail per check
    • Save to previews/phase-8-qa.html
  6. Verdict:

    • PASS → Mark project complete
    • CONCERNS → Note issues, user decides
    • FAIL → Loop back to Phase 7 for fixes (max 3 iterations)

RLHF Loop

After Phase 8 completion:

  1. Ask user for overall rating (1-5)
  2. If approved:
    • Save design decisions to growthOS/design-intelligence/approved/APPROVED-DESIGNS.md
    • Update STYLE-PROFILE.md with learned preferences
  3. If rejected:
    • Save rejection reasons to growthOS/design-intelligence/anti-patterns/REJECTED-DESIGNS.md
    • Update STYLE-PROFILE.md with anti-preferences
  4. The design intelligence base grows with every project.

Quick Reference: File Paths

ResourcePath
Skill definitiongrowthOS/skills/sales-page/SKILL.md
Pipeline state schemagrowthOS/templates/sales-pages/_PIPELINE-STATE.md
Product knowledge schemagrowthOS/voice/offers/_SCHEMA.md
Design doctrinegrowthOS/design-intelligence/DESIGN-DOCTRINE.md
Anti-slop visualgrowthOS/design-intelligence/anti-patterns/AI-SLOP.md
Style RLHF profilegrowthOS/design-intelligence/approved/STYLE-PROFILE.md
Preview servergrowthOS/sales-page-preview/server.py (port 8061)
Project outputgrowthOS/output/sales-pages/{slug}/
Phase previewsgrowthOS/output/sales-pages/{slug}/previews/

Frequently asked questions

What does the Sales Page AI skill do?

Complete 8-phase sales page production system. Orchestrates discovery, research, visual design, narrative, fusion, build, and QA to produce elite conversion-focused landing pages. Integrates Product Knowledge Base, Design Intelligence Base, and Voice Intelligence for directed AI output. Use when creating sales pages, landing pages, or any conversion-focused web page.

Why use Sales Page on TypingMind?

Because you install it once and use it with any model. Sales Page is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Sales Page in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/melgarafael/growthOS/tree/main/skills/sales-page. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Sales Page?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Sales Page?

As many as you like. As long as a model supports skills, you can use Sales Page with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Sales Page AI skill free?

Yes. It is published on GitHub by melgarafael under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇