Agent skill

Karpathy Idea Files

by LearnPrompt in LearnPrompt/andrej-karpathy-skills

Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement.

MITAuto-check passedKnowledge Management

Install Karpathy Idea Files

skills CLI
$ npx skills add LearnPrompt/andrej-karpathy-skills --skill karpathy-idea-files -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install LearnPrompt/andrej-karpathy-skills karpathy-idea-files --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/LearnPrompt/andrej-karpathy-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/karpathy-idea-files .claude/skills/karpathy-idea-files && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
karpathy-idea-files
GitHub stars
110
Token cost
~1.5k tokens
SKILL.md length
176 words
Files
1
Skills in repo
15
Repo updated
First seen
Licence
MIT

At a glance

Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement.

  • The user wants to capture a new idea quickly
  • SKILL.md covers Core Principle, The Idea File Template, Converting a Vague Idea to a… and The Agent Implementation Prompt, plus 6 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Needs to write a project spec for an agent to implement

What it does

Karpathy Idea Files is an agent skill from LearnPrompt/andrej-karpathy-skills. Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement. Use this skill when the user wants to capture a new idea quickly, needs to write a project spec for an agent to implement, wants to share a concept without building it first, or says "write up this idea", "create a spec", "gist this", "idea file", "agent can implement this", "brief for this project". Based on Karpathy 27k-like LLM Wiki Gist post and GitHub repo.

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Knowledge Management, covering LLM wikis. It works with GitHub. The repository describes itself as: Karpathy-inspired Agent Skills collection. The licence is MIT.

When your agent uses it

  • The user wants to capture a new idea quickly
  • Needs to write a project spec for an agent to implement
  • Wants to share a concept without building it first
  • Says write up this idea

Example prompts

  • “write up this idea”
  • “create a spec”
  • “gist this”
  • “/karpathy-idea-files”

What it can do on your machine

Read from SKILL.md and the folder at commit 9e46dec. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    Links to these hosts (documentation or services it may open):

    • x.com
    • github.com

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Karpathy Idea Files loads about 1.5k tokens when it runs. Until then it costs about 123 tokens; SKILL.md has 176 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~123
When it runs · the whole SKILL.md, loaded when a task matches
~1.5k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from LearnPrompt/andrej-karpathy-skills at commit 9e46dec, republished under its MIT licence (© LearnPrompt). 176 words, ~1,482 tokens.

Download SKILL.mdSave it as .claude/skills/karpathy-idea-files/SKILL.md (or your agent's skills folder).
name
karpathy-idea-files
description
Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement. Use this skill when the user wants to capture a new idea quickly, needs to write a project spec for an agent to implement, wants to share a concept without building it first, or says "write up this idea", "create a spec", "gist this", "idea file", "agent can implement this", "brief for this project". Based on Karpathy 27k-like LLM Wiki Gist post and GitHub repo.
disable-model-invocation
false
user-invocable
true
related_skills
karpathy-agentic-engineering, karpathy-minimalism, karpathy-output-evolution, karpathy-llm-wiki

Skill 9: Idea Files / Gist-First Sharing(想法文件优先)

Source: https://x.com/karpathy/status/2040470801506541998 | https://github.com/multica-ai/andrej-karpathy-skills "LLM Knowledge Bases idea file" — 27k likes

Core Principle

Ship the idea, not the implementation.

Karpathy's Gist pattern: write a clear abstract + success criteria + constraints, then let an agent (or anyone) implement it for their specific context. One idea spec → infinite implementations.

The old model: write code → share code → others adapt code. The Karpathy model: write idea spec → share spec → agents implement for context.

The Idea File Template

Every new idea gets this format before any code is written:

markdown
# [IDEA NAME]
> Status: [DRAFT / SPEC / ACTIVE / ARCHIVED]
> Created: [DATE]
> Source inspiration: [LINK or "original"]

## Abstract (2-3 sentences)
What is this? What problem does it solve? Why does it matter?

## The Core Insight
What's the one non-obvious observation that makes this idea work?
(If you can't write this, the idea isn't ready yet.)

## Success Criteria
When is this "done"? List 3-5 verifiable outcomes:
- [ ] [Observable outcome 1]
- [ ] [Observable outcome 2]
- [ ] [Observable outcome 3]

## Minimal Viable Version
Describe the simplest possible thing that would prove the concept works.
(One paragraph. No code yet.)

## Constraints
- Must work with: [tools/stack/dependencies]
- Must NOT require: [what to avoid]
- Scale target: [personal use / team / public]
- Time budget: [hours/days]

## Open Questions
Things I don't know yet that matter:
1. [Question 1]
2. [Question 2]

## Implementation Notes (optional)
Hints for whoever (agent or human) implements this:
- Suggested approach: [rough approach]
- Key risk: [what might go wrong]
- Prior art: [similar things that exist]

## Example Input/Output
Input: [example]
Output: [example]

Converting a Vague Idea to a Spec

When you have a rough idea, use this prompt to crystallize it:

I have a rough idea I want to turn into an implementation-ready spec.

Rough idea: [YOUR VAGUE IDEA IN NATURAL LANGUAGE]

Help me write a proper idea file with:
1. A precise 2-sentence abstract (what + why)
2. The core non-obvious insight (the thing that makes it work)
3. 3-5 verifiable success criteria
4. The minimal viable version (simplest proof of concept)
5. 3 things an implementer needs to know
6. 2 open questions I should resolve before building

Be precise. Remove vagueness. If something is unclear, ask — don't invent.

The Agent Implementation Prompt

Once you have an idea file, hand it to an agent:

Implement this idea spec for my specific context.

Idea spec:
[PASTE IDEA FILE]

My context:
- Environment: [OS, language, tools available]
- Existing codebase: [describe or paste relevant parts]
- Constraints: [anything specific to my setup]

Implementation rules:
1. Follow the success criteria exactly
2. Start with the minimal viable version
3. Ask before adding anything not in the spec
4. Output code + a README explaining how to run it
5. Flag any open questions from the spec that affect implementation

Sharing Idea Files

The Karpathy pattern for public sharing:

Polish this idea file for public sharing as a Gist.

[PASTE IDEA FILE]

Requirements:
- Abstract should be compelling but honest (no hype)
- Success criteria should be so clear anyone could verify them
- Implementation notes should be context-independent
- Add a "Variations" section: 3 different ways someone might implement this differently
- Add a note: "Implementations welcome — tag me in yours"

Format: clean markdown, no implementation code, pure idea.

Idea Capture (Quick Mode)

For capturing ideas on the fly (30 seconds):

I have a quick idea: [ONE SENTENCE]

Extract:
1. Problem solved (1 sentence)
2. Core mechanism (1 sentence)  
3. Minimum viable success criterion (1 testable statement)
4. What I need to figure out before building (top 1-2 questions)

Save as a minimal idea file. I'll flesh it out later.

Idea File Vault Structure

ideas/
├── active/         ← being built right now
│   └── llm-wiki.md
├── backlog/        ← solid specs, not started
│   └── autoresearch.md
├── draft/          ← rough, needs more thinking
│   └── neural-output.md
└── archived/       ← built or abandoned, kept for reference
    └── old-idea.md

Workflow

属于工作流:想法到上线(入口)

位置上游下游
入口(第1步)用户有新想法时直接触发karpathy-agentic-engineering(开始工程化实现)

完整链路:idea-files → agentic-engineering → minimalism → supply-chain-hygiene → vibe-to-agentic

Prompt Contract

text
Turn my idea into an idea file. Produce: 1) Abstract (2-3 sentences, what and why), 2) Target user (who benefits), 3) Success criteria (how to know it's done — verifiable), 4) Constraints (budget, tech stack, non-goals), 5) Implementation sketch (steps an agent could follow), 6) Open questions (what I haven't decided yet). Format as a single Markdown file ready to share as a Gist.

Verification Checklist

  • 想法文件有明确的 Abstract(不是模糊愿景)
  • 成功标准是可验证的(不是「做得好」)
  • 约束条件写了「绝对不做什么」
  • 实现建议足够具体,让不认识你的 Agent 也能开始
  • Open Questions 列出了你还没决定的部分
  • 整个文件可以直接发 Gist,不需要额外解释

© LearnPrompt, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in karpathy-idea-files of LearnPrompt/andrej-karpathy-skills.

Open the folder on GitHubat commit 9e46dec

Compare with similar skills

Karpathy Idea Files next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Karpathy Idea Files compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Karpathy Idea Files this skillLearnPrompt/andrej-karpathy-skills110—~1.5kAutomated safety check: PassMIT
LLM Wiki Upgradesdyckjq-lab/llm-wiki-skill2.5k—~735Automated safety check: NotesNone
Wiki Research Looprohitg00/pro-workflow2.9k—~1.5kAutomated safety check: PassNone
Export TemplateHurricaHjz/second-yourself123—~4.8kAutomated safety check: WarnMIT
Swarmvaultswarmclawai/swarmvault709—~6.1kAutomated safety check: PassMIT
Quartz Wikiliangdabiao/llm-wiki156—~861Automated safety check: PassNone

Similar skills

  • LLM Wiki Upgrade

    sdyckjq-lab/llm-wiki-skill

    升级 llm-wiki 到最新版本。从 GitHub 拉取最新代码并通过官方 install.sh 升级核心主线. An agent skill from sdyckjq-lab/llm-wiki-skill.

    2.5k GitHub stars~735 tokensUpdated 2 mo ago
    Knowledge ManagementAuto-check: notes
  • Wiki Research Loop

    rohitg00/pro-workflow

    Auto-grow a pro-workflow wiki by running a budget-capped BFS research loop over pluggable source fetchers (web, arXiv, GitHub).

    2.9k GitHub stars~1.5k tokensUpdated 11 days ago
    Knowledge ManagementAuto-check passed
  • Export Template

    HurricaHjz/second-yourself

    Sync THIS LLM-Wiki framework with its public GitHub repo, ONE direction per run: --push (vault → repo) publishes your framework; --pull (repo → vault) updates your framework from a newer repo version.

    123 GitHub stars~4.8k tokensUpdated 9 days ago
    Knowledge ManagementAuto-check: warnings
  • Swarmvault

    swarmclawai/swarmvault

    Use SwarmVault when the user needs a local-first knowledge vault that writes durable markdown, graph, search, dashboard, review, chat-session, context-pack, task-ledger, static AI export, retrieval…

    709 GitHub stars~6.1k tokensUpdated 3 mo ago
    Knowledge ManagementAuto-check passed
  • Quartz Wiki

    liangdabiao/llm-wiki

    Set up and deploy Quartz v4 static site for markdown wikis. An agent skill from liangdabiao/llm-wiki.

    156 GitHub stars~861 tokensUpdated 5 mo ago
    Knowledge ManagementAuto-check passed
  • Project Maintainer

    Pratiyush/llm-wiki

    Maintain an llmwiki-style open-source project across the full framework pipeline.

    397 GitHub stars~1.4k tokensUpdated 3 mo ago
    Knowledge ManagementAuto-check passed

More from LearnPrompt/andrej-karpathy-skills

All 15 skills in this repo
  • Karpathy Methodology Index

    LearnPrompt/andrej-karpathy-skills

    Apply Andrej Karpathy AI methodology and principles from his 2023-2026 insights. Use this skill when the user wants to apply Karpathy-style thinking, needs…

    110 GitHub stars~1.5k tokensUpdated 3 mo ago
    Auto-check passed
  • Karpathy Agentic Engineering

    LearnPrompt/andrej-karpathy-skills

    Apply Karpathy-style agentic engineering to any coding or building task.

    110 GitHub stars~1.2k tokensUpdated 3 mo ago
    Auto-check passed
  • AutoResearch Loop

    LearnPrompt/andrej-karpathy-skills

    Sets up an autonomous research loop where an agent runs experiments on git branches, logs results and proposes the next iteration while you approve each hypothesis change.

    110 GitHub stars~1.4k tokensUpdated 3 mo ago
    Auto-check passed
  • Karpathy Education First

    LearnPrompt/andrej-karpathy-skills

    Apply the education-first mindset — make everything you build teachable, create nano-project explanations, write for beginners.

    110 GitHub stars~1.8k tokensUpdated 3 mo ago
    Auto-check passed
  • Karpathy LLM Simulator

    LearnPrompt/andrej-karpathy-skills

    Use LLM as a simulator of expert debates and opposing viewpoints instead of getting a single sycophantic answer.

    110 GitHub stars~1.3k tokensUpdated 3 mo ago
    Auto-check passed
  • Karpathy LLM Wiki

    LearnPrompt/andrej-karpathy-skills

    Build and maintain an LLM-powered personal knowledge base or wiki.

    110 GitHub stars~1.4k tokensUpdated 3 mo ago
    Auto-check passed

Works with

Questions about Karpathy Idea Files

What does Karpathy Idea Files do?

Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement. Karpathy Idea Files is an agent skill from LearnPrompt/andrej-karpathy-skills. Create and share ideas as abstract Gist-style specs instead of code — letting agents or others implement.

When should I use Karpathy Idea Files?

Karpathy Idea Files fits situations like: the user wants to capture a new idea quickly; needs to write a project spec for an agent to implement; wants to share a concept without building it first; says write up this idea.

How do I install Karpathy Idea Files in Claude Code?

Run `npx skills add LearnPrompt/andrej-karpathy-skills --skill karpathy-idea-files -a claude-code`. Or copy the skill folder (karpathy-idea-files in LearnPrompt/andrej-karpathy-skills) into .claude/skills/karpathy-idea-files in your project. Claude Code loads it when a task matches its description.

How do I install Karpathy Idea Files in Codex?

Run `npx skills add LearnPrompt/andrej-karpathy-skills --skill karpathy-idea-files -a codex`. Or copy the skill folder (karpathy-idea-files in LearnPrompt/andrej-karpathy-skills) into .agents/skills/karpathy-idea-files in your project. Codex loads it when a task matches its description.

Can I use Karpathy Idea Files in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add LearnPrompt/andrej-karpathy-skills --skill karpathy-idea-files -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/karpathy-idea-files, .gemini/skills/karpathy-idea-files, .github/skills/karpathy-idea-files and .opencode/skills/karpathy-idea-files in your project.

What does Karpathy Idea Files need to run?

SKILL.md names no scripts, command-line tools or credentials: Karpathy Idea Files is instructions for the agent only.

Does Karpathy Idea Files access the network?

SKILL.md names 2 domains. As links in the text: x.com and github.com. This is read from the text; nothing was executed.

Is Karpathy Idea Files safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Karpathy Idea Files use?

Karpathy Idea Files is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Karpathy Idea Files use?

About 1.5k tokens (SKILL.md is roughly 5.9k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Karpathy Idea Files?

Skills that share tags, products or a category with Karpathy Idea Files: LLM Wiki Upgrade (sdyckjq-lab/llm-wiki-skill, 2.5k stars), Wiki Research Loop (rohitg00/pro-workflow, 2.9k stars), Export Template (HurricaHjz/second-yourself, 123 stars) and Swarmvault (swarmclawai/swarmvault, 709 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Karpathy Idea Files?

LearnPrompt (a GitHub user) maintains it in LearnPrompt/andrej-karpathy-skills, which has 110 GitHub stars. The repository holds 15 skills in this directory. The repository was last updated on July 10, 2026.

Source: LearnPrompt/andrej-karpathy-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.