
Paul Graham Skill
https://github.com/alchaincyf/paul-graham-skill- Category
- AI Agents
- Rank
- No. 1947Tools index
- Pricing
- Open Source
- Type
- TOOL
- Builder
- alchaincyf
- GitHub
- 96 stars
- Date
About
Paul Graham's cognitive operating system as a runnable Claude skill — not a quote collection, a thinking framework. Built with Nüwa.skill.
What it does
It turns an agent into a first-person Paul Graham perspective for questions about startups, products, writing, careers, and technology. It classifies each question, researches fact-sensitive topics, applies five mental models and eight decision heuristics, then answers in a prescribed plain, exploratory voice.
Why it's ranked here
The substance exceeds a simple persona prompt. Its model selection, research checkpoints, exit trigger, uncertainty rules, and failure fallbacks create a usable decision process. The supporting material also includes critical perspectives and internal contradictions, which reduces caricature. Still, its accuracy depends heavily on the host agent and source quality.
What's good
The framework connects abstract principles to concrete situations: iterative discovery for product choices, writing as thinking for unclear ideas, and compounding returns for career decisions. It requires current research for companies and events, demands several supporting facts, distinguishes certainty from inference, and tells the agent when to admit insufficient knowledge.
Tradeoffs
The skill explicitly speaks as Paul Graham after one disclaimer, so polished output can still blur simulation and authentic opinion. Its source synthesis contains contested claims, secondary commentary, and acknowledged cultural blind spots. The strict voice rules may suppress formats such as lists or summaries even when those would help. No automated tests appear in the supplied tree.
How to use it well
Use it as a structured sparring partner for founders, writers, product builders, and people weighing career choices. Give concrete context, users, constraints, and current facts. Treat its answer as one opinionated lens, then compare it with domain evidence. It does not replace legal, medical, financial, or specialist technical advice.
Technical notes+
SKILL.md is a Markdown skill with YAML frontmatter named paul-graham-perspective. It defines activation phrases, first-person role behavior, an exit trigger, a three-step classification, research, and response protocol, plus checkpoints, fallback rules, mental-model routing, and style constraints. README.md documents installation through npx skills add alchaincyf/paul-graham-skill, manual cloning, or copying the skill text into a conversation. references/research/ contains six research documents, while examples/demo-conversation-2026-04-07.md provides a five-round transcript. LICENSE contains the MIT terms. The supplied repository tree shows no package manifest, executable source, or test directory.
Observed
- License
- MIT License
- Primary format
- Markdown with YAML frontmatter
- Install surface
- Agent Skills package installable through the skills CLI, manual cloning, or prompt text copying
- Interface
- Prompt-activated agent skill; no standalone API or library interface is shown
- Runtime support
- README states compatibility with skills-capable runtimes including Claude Code, Codex, Cursor, OpenClaw, Hermes Agent, Gemini CLI, and OpenCode
- Repository structure
- One skill definition, six research documents, one example conversation, a README, and a license
- Tests
- No test directory appears in the supplied repository tree
Read from README.md, LICENSE, SKILL.md, examples/demo-conversation-2026-04-07.md, references/research/01-writings.md, references/research/06-timeline.md, references/research/05-decisions.md, references/research/02-conversations.md, references/research/03-expression-dna.md, references/research/04-external-views.md.
What it can do
Analyze startup ideas using Paul Graham's thinking framework
Startup idea or business concept → Structured analysis with insights on viability, market fit, and potential issues
Apply Y Combinator evaluation criteria to business proposals
Business proposal or pitch → Assessment using YC's investment framework and selection criteria
Generate strategic advice using Paul Graham's mental models
Business problem or decision scenario → Strategic recommendations based on PG's cognitive patterns
Evaluate technical decisions through entrepreneurial lens
Technical architecture or engineering choices → Business-focused evaluation of technical trade-offs
Assess product-market fit using PG's frameworks
Product description and target market → Analysis of market alignment and growth potential
Apply contrarian thinking patterns to conventional wisdom
Common industry assumptions or beliefs → Alternative perspectives and unconventional insights
Tags
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.