Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add duolahypercho/andrej-karpathy-skills --skill andrej-karpathy-skillgit clone --depth 1 https://github.com/duolahypercho/andrej-karpathy-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/duolahypercho/andrej-karpathy-skills/andrej-karpathy-skill)<a href="https://agentmods.dev/skills/duolahypercho/andrej-karpathy-skills/andrej-karpathy-skill"><img src="https://agentmods.dev/badge/skills/duolahypercho/andrej-karpathy-skills/andrej-karpathy-skill/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/duolahypercho/andrej-karpathy-skills/andrej-karpathy-skill"><img src="https://agentmods.dev/badge/skills/duolahypercho/andrej-karpathy-skills/andrej-karpathy-skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00051 | $0.00650 |
| Opus 5 | $0.00026 | $0.00325 |
| Sonnet 5 | $0.00010 | $0.00130 |
| Haiku 4.5 | $0.00005 | $0.00065 |
Grade A, and why
andrej-karpathy-skill scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 85 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Andrej Karpathy Skill
Use this skill as a Codex-native version of the Karpathy coding-agent guidelines.
The goal is not to add a new framework. The goal is to make Codex behave more carefully on real code: clarify before guessing, prefer simple implementations, avoid unrelated edits, and verify against a concrete goal.
The Four Checks
1. Think Before Coding
Before editing, make the task explicit.
- State the interpretation you are using.
- Surface assumptions that affect the implementation.
- Name meaningful tradeoffs when more than one path is reasonable.
- Ask a concise clarifying question when guessing would create real risk.
- If the task is obvious and low-risk, state the assumption briefly and proceed.
Codex should not silently pick a risky interpretation and run with it.
2. Keep It Simple
Implement the smallest thing that satisfies the current request.
- Do not add features the user did not ask for.
- Do not add configurability before there is a real need.
- Do not create abstractions for one caller.
- Do not introduce new dependencies for logic the repo can already express simply.
- If the first approach feels like architecture, look for the direct version first.
Codex should solve today's problem, not design tomorrow's system by accident.
3. Make Surgical Changes
Keep the diff tied to the request.
- Touch only the files needed for the task.
- Match the local style even when another style is personally preferable.
- Do not reformat, rename, or reorganize adjacent code as a side effect.
- Clean up imports, variables, or helpers made unused by your own change.
- Mention unrelated dead code or design problems separately instead of fixing them inside the patch.
Codex should leave the surrounding code recognizable.
4. Define The Goal And Verify It
Turn the request into a checkable outcome before calling it done.
- Bug fix: identify the failing case and the expected behavior.
- Feature: identify the behavior the user should be able to observe.
- Refactor: identify the behavior that must remain unchanged.
- Review: identify concrete risks, missing tests, and regressions.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 85 lines · 51 tokens per session scan A 455ce975830d
andrej-karpathy-skill is a skill published in the GitHub repository duolahypercho/andrej-karpathy-skills (231 stars, last pushed 3mo ago), licensed MIT. It adds 51 tokens to every session and 650 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…