getagentseal/founder-playbook

Founder Playbook is a collection of 15 "skills" (plus one router skill called diagnose) that distill well-known business books into bite-sized, structured reference documents designed for AI assistants like Claude, ChatGPT, or Gemini. Instead of 300-page books full of anecdotes, each skill contains the decision trees, frameworks, scoring checklists, case studies, and fill-in templates from a single book. The goal is not to summarize — it's to give an AI the usable parts of each book so it can actually apply them to a founder's problems.

---

What problem does it solve?

Business books contain useful frameworks, but they're buried in long stories. You read a book, highlight 20 pages, and forget the frameworks six months later. This project extracts the actionable core and formats it so an LLM can pull it up and use it live — for example, when you ask "Should we pivot?" the AI loads The Lean Startup skill and walks you through its Build-Measure-Learn framework.

---

What's inside

Skill	Source book	Use it when...
diagnose	(meta-skill) routes across all 15	You don't know where to start, have multiple problems, or "nothing is working"
mom-test	The Mom Test — Rob Fitzpatrick	Customer interviews, validating ideas without leading questions
four-steps	The Four Steps to the Epiphany — Steve Blank	Finding first customers, Customer Development, Market Type
lean-startup	The Lean Startup — Eric Ries	Build-Measure-Learn, MVPs, pivots, innovation accounting
obviously-awesome	Obviously Awesome — April Dunford	Positioning, category choice, competitive context
crossing-the-chasm	Crossing the Chasm — Geoffrey Moore	Tech adoption, beachhead strategy, mainstream scaling
blue-ocean-strategy	Blue Ocean Strategy — Kim & Mauborgne	Category creation, escaping competition
monetizing-innovation	Monetizing Innovation — Ramanujam & Tacke	Pricing strategy, willingness-to-pay, product + price design
spin-selling	SPIN Selling — Neil Rackham	B2B sales, complex deals, multi-stakeholder selling
100m-offers	$100M Offers — Alex Hormozi	Offer design, packaging, making offers irresistible
100m-leads	$100M Leads — Alex Hormozi	Lead generation, advertising, outbound & inbound strategy
money-models	$100M Money Models — Alex Hormozi	Offer sequencing, upsells, downsells, making customer acquisition self-funding
influence	Influence — Robert Cialdini	Persuasion principles, negotiation, defense against manipulation
traction	Traction — Weinberg & Mares	Growth channels, Bullseye Framework, customer acquisition
storybrand	Building a StoryBrand — Donald Miller	Brand messaging, website copy, email campaigns, SB7 Framework
made-to-stick	Made to Stick — Heath & Heath	Making messages memorable, pitches, presentations, SUCCESs Framework

---

How a skill is structured (progressive disclosure)

Each skill folder contains:

• SKILL.md — entry point (~250–500 lines). Decision trees, key frameworks, quick reference. This is all the AI loads by default.
• frameworks.md — detailed framework breakdowns; loaded when depth is needed.
• cases.md — real company case studies from the book.
• examples.md — templates, worksheets, worked examples.
• integration.md — how this skill relates to (and conflicts with) other skills.

This design means Claude reads only ~100 tokens per skill at startup (just the description), loads ~500 lines when a skill is triggered, and pulls supporting files only on demand.

---

What makes these different from book summaries

• Honest scope statements — each skill says what the book got wrong, what's outdated, and where it doesn't apply.
• Cross-skill conflict resolution — where two frameworks contradict (e.g., Blue Ocean vs. Obviously Awesome on niches), the conflict is named and resolved in integration.md.
• Modern relevance notes — frameworks from the '90s and 2000s are flagged where they don't apply to 2025+ contexts (PLG, AI-native products, privacy regulations).
• They're applied, not just read — structured reference documents with decision trees, scoring rubrics, and fill-in templates.

---

How to install and use

Quick install (Claude Code, etc.):

npx skills add getagentseal/founder-playbook

Manual install: clone the repo and symlink each skill folder into ~/.claude/skills/.

With Claude Code: it auto-triggers based on your question. For example:

• "What's wrong with my startup?" → loads diagnose
• "I'm interviewing customers tomorrow" → loads mom-test
• "Should we pivot?" → loads lean-startup
• "Which marketing channel should we use?" → loads traction
• "How do I price this?" → loads monetizing-innovation

You can also invoke skills directly with slash commands like /lean-startup.

With ChatGPT, Gemini, or any other LLM: open the relevant SKILL.md file and paste it into your conversation. The YAML frontmatter is just metadata — every LLM ignores it and uses the content. For deeper detail, also paste frameworks.md, cases.md, or examples.md.

With Cursor, Windsurf, or Cline: drop the repo into your project's rules or docs folder; the AI references skills when relevant.

---

Recommended sequence (idea → scale)

1. diagnose — don't know where to start? Start here.
2. four-steps — find customers and validate the business model
3. lean-startup — Build-Measure-Learn iteration speed
4. mom-test — how to talk to customers without biasing them
5. obviously-awesome — position the product clearly
6. storybrand — clarify your message so customers listen
7. made-to-stick — make every message memorable and actionable
8. monetizing-innovation — design product + price together
9. 100m-offers — package the offer so it's irresistible
10. money-models — sequence offers so customers fund acquisition
11. spin-selling — close B2B deals
12. crossing-the-chasm — cross from early adopters to mainstream
13. blue-ocean-strategy — escape competition if commoditizing
14. traction — systematic channel selection (Bullseye Framework)
15. 100m-leads — tactical playbook for chosen channels
16. influence — persuasion principles throughout

---

Example use case

Ask your LLM: "I built an AI tool, talked to 30 people who said they love it, but zero have paid. What's wrong?"

Without the skills: generic advice about product-market fit.

With the skills: Claude loads Mom Test (those 30 people were probably being polite — here's what questions to ask instead), Four Steps (you skipped Customer Validation — here's the earlyvangelist pain hierarchy to score them), Lean Startup (your value hypothesis isn't validated — here's innovation accounting to measure real demand), and StoryBrand (your website probably talks about you instead of the customer's problem — here's the BrandScript to fix it).

---

Support the original authors

The README explicitly frames these skills as study aids and critical commentary — they help you apply what the books teach, not replace reading them. It includes a table with links to buy each of the 15 books.

License

MIT.