The Founder Decision Engine
A decision operating system for startup founders, powered by Claude Code. Synthesizes the wisdom of 25+ legendary founders, investors, and thinkers into an interactive AI advisor that helps you make the highest-leverage decisions at every stage of building a company.
What This Is
This is not a reading list. It's a decision engine — an interactive system that:
- Routes your decisions through battle-tested frameworks from the world's best founders
- Challenges your assumptions before you commit
- Runs pre-mortems on your plans before they fail
- Evaluates your startup through investor eyes
- Keeps you honest with weekly and quarterly checklists
It runs as a Claude Code project. When you boot it up, Claude becomes a strategic co-founder who has internalized the decision-making of Paul Graham, Steve Jobs, Jeff Bezos, Naval Ravikant, Warren Buffett, Peter Thiel, Elon Musk, Marc Andreessen, and 15+ others.
Quick Start
cd ~/decision-engine claude
That's it. The CLAUDE.md kernel loads automatically and transforms Claude into your founder advisor. Then use slash commands:
/decide should I pursue enterprise or SMB customers first?
/build an AI tool that automates bookkeeping for freelancers
/pivot we've been at this for 8 months with 200 users but no growth
/hire considering bringing on a VP of Sales at Series A
/raise thinking about raising a $3M seed at $15M post
/contrarian a marketplace connecting local chefs with home diners
/premortem launching in 3 new cities simultaneously next quarter
/weekly
Slash Commands
| Command | Purpose | What It Does |
|---|---|---|
/decide <decision> |
Universal router | Classifies your decision, runs the 5-question Pre-Decision Protocol, applies the right frameworks, gives a recommendation |
/build <idea> |
Idea validation | Runs Graham's Three-Filter, Thiel's Contrarian Test, Caldwell's Tar Pit Check, Sequoia's PMF Archetypes, and Chesky's 11-Star design |
/pivot <context> |
Pivot vs. persevere | Diagnostic on time, user signals, hypothesis quality, and energy direction. Applies YC's pivot matrix and Buffett's punch card |
/hire <role/context> |
Hiring decisions | Pre-hire necessity check, candidate evaluation (Buffett's integrity test, Musk's problem-solving interview, Rabois's barrel classification) |
/raise <context> |
Fundraising strategy | Default alive check, readiness checklist, amount formula, investor selection criteria, process playbook, term sheet red flags |
/scale <context> |
Scaling diagnostic | PMF reality check, a16z phase identification, blitzscaling criteria, executive hiring sequence, danger signals |
/wartime <context> |
Leadership mode | Diagnoses peacetime vs. wartime, provides mode-specific operating playbook, identifies switch triggers |
/weekly |
Weekly check-in | Interactive 10-question founder checklist covering metrics, focus, customers, speed, and long-term thinking |
/contrarian <idea> |
Stress test | Thiel's contrarian framework, consensus check, Graham's "sounds bad/actually good" filter, Dixon's timing signal, Buffett's inversion |
/premortem <plan> |
Failure analysis | Assumes the plan failed. Generates 5+ failure scenarios, runs Munger's bias audit, elimination round, Buffett scorecard, clear verdict |
Agents
Specialized sub-agents for deeper analysis. Claude Code can invoke these automatically or you can request them.
| Agent | Role | Voice |
|---|---|---|
| advisor | Full-context strategic advisor | Synthesized — all 25+ thinkers |
| contrarian | Devil's advocate that challenges every assumption | Peter Thiel + Charlie Munger |
| premortem | Inversion analyst — assumes failure, finds kill shots | Charlie Munger + Warren Buffett |
| investor-lens | Evaluates from top-tier VC perspective | Marc Andreessen + Sequoia Capital |
Example Usage
Ask Claude to invoke agents directly:
"Run the contrarian agent on my idea for an AI writing assistant"
"Have the investor-lens agent evaluate my startup as if I'm pitching at a partner meeting"
"I need a pre-mortem on our decision to go all-in on enterprise sales"
Knowledge Base
The decision engine is built on 6 research files totaling 185K+ of distilled, actionable frameworks:
| File | Lines | Contents |
|---|---|---|
00-DECISION-ENGINE.md |
586 | Start here. Master index, 7 meta-frameworks, 5 decision playbooks (Build, Pivot, Raise, Scale, Wartime), weekly/quarterly checklists, one-page cheat sheet |
01-yc-wisdom.md |
671 | Paul Graham (6 frameworks), Sam Altman (execution stack), Michael Seibel (MVP/pivot), Dalton Caldwell (tar pits), Jessica Livingston (founder character). 40+ quotes, 5 decision trees, 5 quick checklists |
02-vc-wisdom.md |
961 | Marc Andreessen (BPMF/APMF, idea maze), Ben Horowitz (peacetime/wartime, management debt), Keith Rabois (barrels vs. ammunition), Elad Gil (scaling playbook), Chris Dixon (timing signals), Sequoia (3 PMF archetypes), Vinod Khosla (black swan investing). 16 fundraising commandments |
03-jobs-gates.md |
429 | Steve Jobs (350→10 elimination, product intuition, A-player compounding, NeXT lessons) and Bill Gates (Think Weeks, IBM deal, Internet Tidal Wave, platform thinking). 10 actionable frameworks, comparison table |
04-naval-buffett-ellison.md |
511 | Naval Ravikant (leverage hierarchy, specific knowledge, 4 types of luck), Warren Buffett + Charlie Munger (circle of competence, moats, inversion, punch card, 8 mental models), Larry Ellison (zero-sum strategy, acquisition framework, platform lock-in). 5 cross-cutting frameworks |
05-modern-founders.md |
689 | Elon Musk (first principles, 5-step algorithm, idiot index), Jeff Bezos (Day 1, two-way doors, flywheel, regret minimization), Peter Thiel (Zero to One, contrarian framework, power law), Reid Hoffman (blitzscaling), Brian Chesky (11-star experience), Patrick Collison (talent density), Jensen Huang (platform bets). 10 actionable frameworks |
The Meta-Frameworks
Seven cross-cutting principles that ALL the greats share, extracted from the full research:
-
Focus Is Subtraction — Jobs (350→10), Musk (delete algorithm), Bezos (Day 1), Chesky (COVID cuts), Buffett ("say no to almost everything")
-
The Leverage Hierarchy — Code/media > capital > labor > your time. Always push toward higher leverage. (Naval)
-
The Moat Imperative — Network effects > switching costs > tech > brand > scale. If a billionaire could destroy you, fix that first. (Buffett, Thiel, a16z)
-
The Speed Doctrine — 70% info is enough for reversible decisions. "If the schedule is long, it's wrong." (Bezos, Musk, Collison, Altman)
-
The People Equation — First 5 hires set the company's permanent ceiling. Barrels > ammunition. Integrity is non-negotiable. (Jobs, Rabois, Buffett)
-
The Contrarian Imperative — If everyone agrees, the opportunity is gone. The best ideas look wrong to most smart people. (Thiel, Graham, Buffett, Dixon)
-
The Compounding Principle — Always choose the option that builds value over time. Short-term thinkers are your competition. (Naval, Buffett, Bezos, Huang)
The One-Page Cheat Sheet
╔═══════════════════════════════════════════════════════════╗
║ HIGHEST LEVERAGE DECISION RULES ║
╠═══════════════════════════════════════════════════════════╣
║ ║
║ 1. SUBTRACT FIRST (Jobs, Musk, Bezos) ║
║ → What should NOT exist? Delete it. ║
║ ║
║ 2. SPEED > PERFECTION (Bezos, Musk, Collison) ║
║ → 70% info is enough. Decide. Iterate. ║
║ ║
║ 3. SEEK ASYMMETRY (Naval, Buffett, Thiel) ║
║ → Capped downside + unlimited upside = go. ║
║ ║
║ 4. BUILD MOATS (Buffett, Thiel, Ellison) ║
║ → If a billionaire could destroy you, fix that. ║
║ ║
║ 5. HIRE BARRELS (Rabois, Jobs, Musk) ║
║ → One barrel > ten ammunition. ║
║ ║
║ 6. COMPOUND EVERYTHING (Naval, Buffett, Bezos) ║
║ → Choose the option that builds over time. ║
║ ║
║ 7. THINK CONTRARIAN (Thiel, Graham, Buffett) ║
║ → If everyone agrees, the opportunity is gone. ║
║ ║
║ 8. OBSESS OVER CUSTOMERS (Bezos, Graham, Chesky) ║
║ → Not competitors. Not metrics. Customers. ║
║ ║
║ 9. INVERT THE QUESTION (Munger, Buffett) ║
║ → "How does this fail?" before "How does this win?" ║
║ ║
║ 10. PLAY THE LONG GAME (Bezos, Naval, Huang) ║
║ → Short-term thinkers are your competition. ║
║ ║
╚═══════════════════════════════════════════════════════════╝
Architecture
decision-engine/
│
├── CLAUDE.md # Kernel — loaded into every session
│ # Defines identity, principles, behavior
│
├── 00-DECISION-ENGINE.md # Master index + playbooks + checklists
├── 01-yc-wisdom.md # Knowledge base: YCombinator
├── 02-vc-wisdom.md # Knowledge base: VCs & a16z
├── 03-jobs-gates.md # Knowledge base: Jobs & Gates
├── 04-naval-buffett-ellison.md # Knowledge base: Naval, Buffett, Ellison
├── 05-modern-founders.md # Knowledge base: Musk, Bezos, Thiel, etc.
│
└── .claude/
├── commands/ # Slash commands (interactive playbooks)
│ ├── decide.md # /decide — universal router
│ ├── build.md # /build — idea validation
│ ├── pivot.md # /pivot — pivot vs. persevere
│ ├── hire.md # /hire — hiring framework
│ ├── raise.md # /raise — fundraising strategy
│ ├── scale.md # /scale — scaling diagnostic
│ ├── wartime.md # /wartime — leadership mode
│ ├── weekly.md # /weekly — founder checklist
│ ├── contrarian.md # /contrarian — stress test
│ └── premortem.md # /premortem — failure analysis
│
├── agents/ # Specialized sub-agents
│ ├── advisor.md # General founder advisor
│ ├── contrarian.md # Devil's advocate
│ ├── premortem.md # Inversion analyst
│ └── investor-lens.md # VC perspective evaluator
│
└── skills/ # Lazy-loaded capabilities
├── decision-frameworks/
│ └── SKILL.md # Auto-activates on decision questions
└── founder-wisdom/
└── SKILL.md # Auto-activates for specific thinkers
How It Works
| Layer | Role | Analogy |
|---|---|---|
CLAUDE.md |
Core identity and operating principles | OS kernel |
| Knowledge base (00-05) | Decision frameworks and founder wisdom | System libraries |
| Slash commands | Interactive decision playbooks | CLI tools / system calls |
| Agents | Specialized analytical perspectives | Co-processors |
| Skills | Context-aware framework activation | Drivers / modules |
Thinkers Included
Founders
| Name | Company | Primary Contribution |
|---|---|---|
| Steve Jobs | Apple, Pixar, NeXT | Focus, product intuition, A-player teams |
| Bill Gates | Microsoft | Platform thinking, competitive strategy, Think Weeks |
| Elon Musk | Tesla, SpaceX | First principles, speed, the deletion algorithm |
| Jeff Bezos | Amazon | Day 1 thinking, two-way doors, customer obsession |
| Larry Ellison | Oracle | Competitive strategy, acquisitions, platform lock-in |
| Brian Chesky | Airbnb | 11-star experience, doing things that don't scale |
| Patrick Collison | Stripe | Infrastructure thinking, talent density, craft |
| Jensen Huang | NVIDIA | Platform bets, AI as infrastructure |
Investors & Thinkers
| Name | Affiliation | Primary Contribution |
|---|---|---|
| Paul Graham | Y Combinator | Idea validation, PMF, "Do Things That Don't Scale" |
| Sam Altman | Y Combinator, OpenAI | Execution speed, team building |
| Marc Andreessen | a16z | Market > Product > Team, PMF, scaling |
| Ben Horowitz | a16z | Peacetime/wartime CEO, management debt |
| Peter Thiel | Founders Fund | Zero to One, contrarian thinking, monopoly strategy |
| Naval Ravikant | AngelList | Leverage, specific knowledge, wealth creation |
| Warren Buffett | Berkshire Hathaway | Moats, circle of competence, long-term compounding |
| Charlie Munger | Berkshire Hathaway | Inversion, mental models, bias detection |
| Reid Hoffman | LinkedIn, Greylock | Blitzscaling, network effects |
| Keith Rabois | Khosla, Founders Fund | Barrels vs. ammunition, operations |
| Chris Dixon | a16z | Timing signals, idea maze |
| Elad Gil | Color Genomics | Post-PMF scaling, executive hiring |
| Vinod Khosla | Khosla Ventures | Black swan investing, founder criteria |
| Michael Seibel | Y Combinator | MVP philosophy, pivot vs. iterate |
| Dalton Caldwell | Y Combinator | Tar pit ideas, pivot quality |
| Jessica Livingston | Y Combinator | Founder character assessment |
Institutions
| Name | Primary Contribution |
|---|---|
| Sequoia Capital | 3 PMF archetypes (Hair on Fire, Hard Fact, Future Vision) |
| a16z | Network effects taxonomy, fundraising commandments, scaling phases |
| Y Combinator | Startup school curriculum, batch wisdom from 1,000+ startups |
Sources
Research was conducted across 50+ web sources including:
- Paul Graham's essays (paulgraham.com)
- Sam Altman's Startup Playbook
- a16z blog and podcast archives
- Jeff Bezos's annual shareholder letters (1997-2021)
- Peter Thiel's CS183 Stanford lectures and Zero to One
- Walter Isaacson's biographies of Steve Jobs and Elon Musk
- Ben Horowitz's The Hard Thing About Hard Things
- The Almanack of Naval Ravikant
- Reid Hoffman's Blitzscaling
- Elad Gil's High Growth Handbook
- Warren Buffett's annual letters to Berkshire Hathaway shareholders
- Charlie Munger's Poor Charlie's Almanack
- Sequoia Capital's Arc PMF framework
- YCombinator's seed fundraising guide
- Farnam Street (fs.blog) interview transcripts
- Masters of Scale podcast transcripts
- Hacker News founder wisdom threads
- D Conference and Lost Interview transcripts
Extending the Engine
Add a New Thinker
- Create or extend a knowledge base file (01-05) with their frameworks
- Add them to the
founder-wisdomskill in.claude/skills/founder-wisdom/SKILL.md - Update the source map in
00-DECISION-ENGINE.md
Add a New Command
- Create a new
.mdfile in.claude/commands/ - Use
$ARGUMENTSfor user input - Reference the relevant knowledge base files
- Add it to
CLAUDE.mdand this README
Add a New Agent
- Create a new
.mdfile in.claude/agents/with YAML frontmatter - Define
name,description, andallowedTools - Give it a clear voice, framework set, and response format
- Add it to
CLAUDE.mdand this README
License
This is a personal decision tool. The frameworks, quotes, and mental models referenced herein are attributed to their original creators. This project is a synthesis for personal use and education.
Built with Claude Code agent teams. 4,835+ lines of synthesized founder wisdom. March 2026.