Convert Anthropic's AI-native startup playbook into a reusable agent skill. Validates your startup idea, builds your MVP, launches your product, and scales your company — all guided by your AI agent. Works with Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, OpenClaw, and 50+ agents.
42% of startups fail because they build something nobody wants. In 2026, agentic coding makes it faster to build a product nobody wants than ever before.
This skill gives you the discipline of the framework without losing the speed of AI. It's adapted from Anthropic's official "Founder's Playbook: Building an AI-Native Startup" — with all platform-specific product references replaced so any agent, on any platform can use it.
# Claude Code, Cursor, Codex, Copilot, Gemini CLI, etc.
npx skills add lloyd-c137/the-founders-playbook
# OpenClaw
npx clawhub install the-founders-playbookThat's it. Your agent now has the full startup lifecycle framework.
Don't write a line of code until you have evidence. This stage runs you through:
- Problem hypothesis pressure testing with adversarial analysis
- 4-layer competitive landscape mapping (direct, indirect, acquirer, adjacent)
- Customer discovery interview protocol design
- TAM/SAM/SOM modeling with built-in skepticism
- Exit criteria: Problem-solution fit confirmed
Turn your validated idea into a real product without drowning in AI-generated tech debt:
- Architectural context file creation (before a line of code)
- Scope definition (what's in, what's NOT in, what evidence justifies adding)
- Pre-launch security review (auth, injection, data exposure, dependencies)
- Measurement framework with Sean Ellis test and retention benchmarks
- Pivot decision framework (when to adjust vs. when to start over)
- Exit criteria: Genuine product-market fit signals
Turn early traction into a repeatable growth engine:
- Founder bottleneck audit (what breaks if you're unavailable for a week?)
- Technical debt sprint planning by impact
- Compliance baseline (SOC2, GDPR, HIPAA mapping)
- GTM engine (market segmentation, messaging architecture, sales playbook)
- Fundraising prep (investor memo, financial model, data room)
- Exit criteria: Repeatable revenue + operations that don't need you daily
Turn your product into a durable company with defensible moats:
- Enterprise infrastructure gap analysis (SLAs, security reviews, procurement)
- Data moat analysis (behavioral patterns → feedback loop → competitive advantage)
- Workflow integration audit (switching cost analysis per customer)
- Domain expertise → AI context conversion (proprietary knowledge substrate)
- Exit criteria: Institutional infrastructure + enterprise contracts
| Other startup advice | This skill | |
|---|---|---|
| AI-native | Generic advice about building | Built for the 2026 reality of agentic coding |
| Stage-aware | One-size-fits-all | 4 stages with transition detection |
| Actionable | "Talk to users" | Full interview protocol, analysis framework |
| Platform-agnostic | Tied to one AI product | Any agent, any coding assistant |
| Exercise-driven | Theory | 18 concrete exercises you run with AI |
Tell your AI agent where you are in natural language — the skill detects your stage automatically:
| You Say | Stage Activates |
|---|---|
| "I have an idea for..." | Idea — validation, research, customer discovery |
| "Help me build the MVP" | MVP — architecture, scope, coding, security |
| "We're ready to launch" | Launch — GTM, compliance, fundraising |
| "We need to scale" | Scale — enterprise, moats, infrastructure |
When you hit a stage's exit criteria, your agent suggests transitioning to the next stage.
├── README.md ← This file
├── LICENSE ← MIT-0 (free, no attribution required)
├── CONTRIBUTING.md ← How to contribute
├── AGENTS.md ← Agent workspace guidelines
└── skills/the-founders-playbook/
├── SKILL.md ← Complete playbook (4 stages, 18 exercises)
├── README.md ← Skill documentation
├── setup.md ← First-use guidance
└── references/
└── stage-reference.md ← Quick reference cheat sheet
- Solo founders and one-person companies building with AI
- Technical founders who need business/distribution counterweight
- Non-technical founders who want a structured build framework
- Indie hackers and solopreneurs shipping products with agentic coding
- Anyone who read Anthropic's playbook and wants it as an executable skill
MIT-0 — Free to use, modify, redistribute. No attribution required. Building on Anthropic's original work with love for the entire agent ecosystem.
Found a bug? Want to contribute? Open an issue or submit a PR. All contributions welcome.