Repository Details
Shared by
HelloGitHub Rating
0 ratings
Behavior Norms to Minimize Errors in AI Coding Assistants
Past 2 days Received 3.4k stars ✨
Claim
Discuss
Collect
Share
93.8k
Stars
No
Chinese
Other
Language
Yes
Active
1
Contributors
74
Issues
No
Organization
None
Latest
9k
Forks
None
License
More
This project is inspired by Andrej Karpathy's observations on LLM coding pitfalls. It refines four principles: thinking before coding, prioritizing conciseness, precise modification, and verifying target results, and encapsulates them into a skill pack. It supports one-click installation into tools like Claude Code and Cursor, effectively reducing issues such as AI randomly modifying code, over-abstracting, and guessing without asking.
Comments
Rating:
No comments yet