Are you the author? Sign in to claim
An open-source AI skill that upgrades your prompts. Load verified thinking profiles of tech founders and philosophers to

Borrowed Brain Pro is powered by Two Core Superpowers:
- Distill Engine: Research and distill ANY public figure (past or present) into a reusable, structured Thinking Profile (
profiles/name.md).- Decision Engine: Apply pre-built or custom profiles (single lens or multi-lens contrast) to your real decision dilemmas.
| Approach | Prompt Example | Output Quality | Value for Real Decisions |
|---|---|---|---|
| Generic AI Prompt | "Think like Warren Buffett and advise me" | Generic motivational quotes ("Focus on long-term value") | Zero — Unverifiable, cliché summary |
| Borrowed Brain Pro | "Apply Buffett's profile to my SaaS expansion deal" | Layered research, sourced moments & explicit failure boundaries | High — Actionable decision lens with concrete limits |
flowchart TD
User(["User Decision / Query"]) --> Router{"Mode Router"}
Router -->|1. Distill Mode| Distill["4-Layer Web Research\n• Primary Voice & Interviews\n• Documented Decisions\n• Third-Party Biographies\n• Criticisms & Failures"]
Distill --> Save["Save Profile\nprofiles/name.md"]
Router -->|2. Apply Mode| Apply["Single Lens Reasoning\nApply 1 Profile → Surface Hidden Blind Spots"]
Router -->|3. Compare Mode| Compare["Multi-Lens Contrast\nCompare 2+ Profiles → Map Conflicts & Gaps"]
Router -->|4. Boardroom Mode| Boardroom["Virtual Board Debate\nConvene 3-4 Figures → Cross-Examine & Output Consensus"]
Save --> Apply
Save --> Compare
Save --> Boardroom
Apply --> Output["Decision Lens Output\n(Explicitly flags what the lens misses)"]
Compare --> Output
Boardroom --> Output
# 1. Distill a new figure
"Build a thinking profile for Charlie Munger"
# 2. Apply a single profile to your situation
"Using Munger's profile, should I expand into this new market?"
# 3. Compare two profiles on a dilemma
"Compare Jobs and Altman's profiles on my launch timing dilemma"
# 4. Convene a 4-Figure Virtual Boardroom Debate
"Convene the Boardroom (Jobs, Munger, Altman, Hastings) on my pricing dilemma"
Simply install the Skill and say: "Convene the Boardroom on [your dilemma]". The AI automatically launches a 3-round multi-figure debate:
🏛️ VIRTUAL BOARDROOM DEBATE
Topic: "Should we lower our SaaS prices to capture market share?"
Board Members: Steve Jobs, Charlie Munger, Sam Altman, Reed Hastings
Round 1 — Opening Stances
- Munger: "Margin reduction increases risk. Invert the problem: why isn't the product valuable enough at full price?"
- Jobs: "Never compete on price. If you discount, you're admitting lack of quality."
- Altman: "Ship fast, iterate pricing dynamically. Scale users first, monetize later."
- Hastings: "Farm for dissent internally before changing prices. What do your sales reps privately say?"
Round 2 — Direct Cross-Examination
- Munger → Altman: "Your obsession with raw user growth ignores unit economics."
- Jobs → Hastings: "Seeking consensus with sales reps leads to mediocrity. Focus on product perfection."
Round 3 — Consensus & Irreconcilable Friction
- Unanimous Consent: Do not cut price without cutting non-essential features.
- Core Friction: Jobs (never discount) vs. Altman (scale first).
Executive Verdict & Blind Spot Warning:
- Action Plan: Maintain core price tier, launch a stripped-down entry tier.
- Board Blind Spot: None of these founders ran low-margin B2B SMB software.
Example 2: Philosophy & Engineering Convergence
🏛️ VIRTUAL BOARDROOM DEBATE
Topic: "Should we implement a strict top-down engineering architecture or let teams organically decide their own tech stack?"
Board Members: Linus Torvalds, Laozi, Socrates
Round 1 — Opening Stances
- Torvalds: "Top-down architecture is garbage. Give them the absolute minimum constraints (like git's core data structure) and let them build what works."
- Laozi: "Wu Wei. Force breeds resistance. The best architecture is like water, flowing into the natural shape of the organization. Decentralize."
- Socrates: "Before deciding, what is your definition of 'strict'? Have you examined the false certainty that a single tech stack will solve your actual product problems?"
Round 2 — Direct Cross-Examination
- Socrates → Torvalds: "You say 'minimum constraints'. Who defines the minimum? If you dictate the minimum, are you not still a top-down architect?"
- Torvalds → Laozi: "Being like water is fine until the code doesn't compile. Water without a container is just a puddle. You still need good taste in data structures."
Round 3 — Consensus & Irreconcilable Friction
- Unanimous Consent: Heavy-handed top-down mandates will fail and cause engineering revolt.
- Core Friction: Torvalds (needs rigid low-level data structures) vs. Laozi (pure organic emergence).
Executive Verdict & Blind Spot Warning:
- Action Plan: Define only the API boundaries and core data structures. Let individual teams choose their internal tech stack.
- Board Blind Spot: This ignores enterprise compliance and security audit requirements which often mandate centralized tech stacks.
Standard AI regurgitates success quotes. Borrowed Brain Pro audits blood-and-tears failures. We study documented crises so you don't repeat them:
| Case File | Figure | The Disaster | Key Principle Extracted |
|---|---|---|---|
| Munger's Alibaba Trap | Charlie Munger | 70% stock crash on $100M+ BABA bet | Low valuation cannot save an eroding tech moat; retail loyalty in tech decays fast. |
| Hastings' Qwikster Split | Reed Hastings | 800k lost subscribers & 77% stock drop | Never let internal org chart convenience force customer UX friction. Farm for dissent. |
| Jobs' NeXT Workstation Flop | Steve Jobs | $12,000 price tag, 50k sales in 8 years | Aesthetic perfection cannot compensate for price ceiling blindness. |
| Musk's Model 3 Hell | Elon Musk | Stalled Fremont factory & sleeping under desk | Never automate a process step before deleting and simplifying it first. |
Run a Failure Audit: Simply ask
"Run a Failure Audit on Munger's Alibaba investment"to inspect the full post-mortem.
Don't know which individual profiles to pick? Load a pre-configured Specialist Advisory Board:
| Pack | Focus Domain | Board Members Included | Quick Command |
|---|---|---|---|
| Startup Founders Pack | Product, Speed & Scaling | Jobs, Altman, Musk, Huang, Graham, Hastings | "Convene the Startup Founders Pack on [dilemma]" |
| Investors Pack | Capital & Valuation | Buffett, Munger, Voss | "Convene the Investors Pack on [deal dilemma]" |
| Engineering Pack | System Design & Code Taste | Torvalds, Musk, Feynman, Evans, Newport | "Convene the Engineering Pack on [architecture dilemma]" |
| Crisis Leadership Pack | Crisis & Mental Fortitude | Aurelius, Munger, Feynman, Newport | "Convene the Crisis Leadership Pack on [crisis dilemma]" |
No walls of text. Every core profile includes a native Mermaid Decision Flowchart rendering instantly on GitHub:
flowchart TD
Deal(["Warren Buffett Deal Filter"]) --> CheckCompetence{"Within Circle of Competence?"}
CheckCompetence -->|No| Reject1["❌ REJECT: Outside Circle of Competence"]
CheckCompetence -->|Yes| CheckMoat{"Durable Competitive Moat & Pricing Power?"}
CheckMoat -->|No| Reject2["❌ REJECT: Erodible Moat / Commodity Business"]
CheckMoat -->|Yes| CheckMargin{"Margin of Safety (Price vs. Value)?"}
CheckMargin -->|No| Wait["⏳ WAIT: Fair Business, Unfavorable Price"]
CheckMargin -->|Yes| Execute["✅ EXECUTE: High-Conviction Investment"]
How do we prove our profiles accurately model real decision-making? We run reverse blind-tests on historical crises before the outcome was known:
| Crisis Benchmark | Figure | Historical Context | Accuracy Score |
|---|---|---|---|
| 2008 Financial Crisis | Warren Buffett | Sept 2008 Lehman collapse & Wall St liquidity freeze | 98% Match (Predicted exact 10% preferred stock + warrant structure) |
| 1997 Apple Bankruptcy | Steve Jobs | Aug 1997 Apple 90 days from insolvency | 100% Match (Predicted exact 70% product cut & 2x2 grid) |
| 2023 OpenAI Governance Crisis | Sam Altman | Nov 2023 Board firing & reinstatement | 95% Match (Predicted rapid external leverage & 5-day return) |
| 2006 Nvidia CUDA Crisis | Jensen Huang | Nov 2006 50% stock drop during CUDA multi-billion bet | 96% Match (Predicted unwavering platform conviction over short-term margins) |
Run a Benchmark Eval: Simply ask
"Run a Blind Eval on Buffett's 2008 liquidity decision"to verify profile predictive accuracy.
Every principle in every profile is structured with rigorous epistemic boundaries:
Principle: Actively solicit disagreement rather than assuming silence means agreement ("farming for dissent").
Where it shows up: Built directly out of the Qwikster failure (2011) — Hastings learned that multiple VPs privately doubted the split but didn't say so.
Where it likely breaks down: The principle was reactive, not proactive — instituted AFTER an 800k subscriber drop and 77% stock collapse.
| Person | Domain | Core Strength / Best For |
|---|---|---|
| Warren Buffett | Investing | Opportunity evaluation, valuation discipline, knowing when to say no |
| Charlie Munger | Multidisciplinary | Inverting complex problems, mental models, avoiding stupidity & bias |
| Steve Jobs | Product / Vision | Feature cuts, launch timing, aggressive simplification, relentless quality |
| Chris Voss | Negotiation | Breaking deadlocks, contract negotiations, handling hostile counterparts |
| Richard Feynman | Research / Reasoning | Auditing logic, hunting for self-deception, explaining complex ideas simply |
| Cal Newport | Productivity | Focus protection, structural fixes over willpower, deep work systems |
| Reed Hastings | Leadership | Farming for dissent, building candor, executing major strategic pivots |
| Travis Kalanick | Aggressive Growth | Market entry under resistance, speed vs. compliance trade-offs |
| Julia Evans | Technical Writing | Explaining hard topics simply, writing for past self, clear documentation |
| Sam Altman | AI & Startups | Iterative deployment, speed vs. safety, framing controversial decisions |
| Paul Graham | Startups & Essays | Doing things that don't scale, founder earnestness, keeping identity small |
| Linus Torvalds | Open Source / Architecture | Pragmatic software architecture, code simplicity, managing decentralized contributors |
| Elon Musk | Deep Tech / Engineering | First principles physics reduction, 5-step engineering algorithm, deletion of requirements |
| Jensen Huang | Hardware / Platform | Betting on zero-billion-dollar markets, full-stack co-design, flat organizational execution |
| Marcus Aurelius | Philosophy / Crisis | Stoic leadership under crisis, controlling internal response vs external chaos, duty |
| Friedrich Nietzsche | Philosophy | Amor Fati, antifragility through adversity, self-overcoming |
| Socrates | Philosophy | Epistemic humility, Socratic questioning, auditing false certainty |
| Laozi | Philosophy / Management | Wu Wei (effortless action), water dynamics, organic emergence |
| Platform | Recommended Installation Method | One-Click Setup Instructions |
|---|---|---|
| ChatGPT / Custom GPT | Single-File System Prompt | Copy & paste full text of borrowed-brain-bundle.md into System Prompt |
| Cursor / Windsurf AI | Native .cursorrules | Clone repo or copy .cursorrules into project root |
| Ollama / Local LLMs | Modelfile / System Prompt | Copy & paste borrowed-brain-bundle.md into Modelfile system field |
| Claude / CLI | Git Clone (Skill) | git clone https://github.com/DOTfei/borrowed-brain-pro.git ~/.claude/skills/borrowed-brain-pro |
| Claude.ai | Skill / System Prompt | Upload SKILL.md via Settings → Skills, or paste borrowed-brain-bundle.md |
| Open WebUI | System Prompt | Paste borrowed-brain-bundle.md into System Prompt field |
[!IMPORTANT]
- No Fabricated Quotes: Direct quotes stay under 15 words per source. Everything else is paraphrased.
- Traceable Claims: Every claim must trace to a verified, documented public event or primary source.
- Public Figures Only: Refuses to build profiles for private individuals to protect privacy.
- Limits Always Flagged: Every response explicitly closes with "Here is what this framework misses in your specific situation."
Contributions, source improvements, and new distilled profiles are welcome! See CONTRIBUTING.md for guidelines.
# Automated Bundle Build (Runs automatically via GitHub Actions)
python scripts/build_bundle.py
Design enforcement with memory — keeps your UI consistent across a project
Create animation-rich HTML presentations using a coding agent's frontend skills
1000+ skills curated from Anthropic, Vercel, Stripe, and other engineering teams
Detects 37 AI writing patterns and rewrites text with human rhythm across 5 voice profiles