Progress will not be saved. This browser is blocking storage (often private browsing). You can still learn, but closing the tab loses your progress.
multica-ai/andrej-karpathy-skills
A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls.
- Stars
- 234,927
- +7 days
- —
- Forks
- 23,313
- Language
- —
- License
- —
- Last push
- 20 Apr 2026
Summary
Summary written by AI from the README; it may be imprecise.This project provides behavioral guidelines for Claude Code formatted as a CLAUDE.md file, a Claude Code plugin, or a Cursor rule. Derived from Andrej Karpathy's observations on LLM coding pitfalls, it instructs the model to clarify assumptions upfront, avoid overengineering, make surgical edits, and verify tasks with test loops.
Use it when
- When Claude Code overengineers solutions or introduces unnecessary abstractions
- When you want Claude Code to touch only relevant lines and avoid unsolicited refactoring
- When setting up test-driven, goal-oriented verification loops for assistant tasks
Install
/plugin marketplace add forrestchang/andrej-karpathy-skills/plugin install andrej-karpathy-skills@karpathy-skillsCommands quoted from the repo's README. Read the README before running them.
GitHub shows no license for this repo. Ask the author before using it at work.
Stars over 90 days
Collecting: 2 of 90 days so far. The chart appears after 7 days.