Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
57 lines (44 loc) · 2.28 KB

CONTRIBUTING.md

File metadata and controls

57 lines (44 loc) · 2.28 KB
Raw

Contributing to AI Security Mastery

Thanks for reading closely enough to want to improve this. Contributions that make the book more correct, clearer, or more current are welcome.

What's welcome

  • Corrections: technical errors, broken code, wrong math, stale claims. These are the most valuable contributions a book can receive.
  • Clarity fixes: a passage that confused you probably confuses others. Say where you got stuck and why.
  • Currency: AI security moves fast. If an attack, defense, or reference in the book has been superseded, open an issue with a source.
  • Suggestions for v2.0: especially in agentic/MCP security — see the Roadmap in the README.

What's not a fit

  • Large unsolicited content drops (whole new sections or chapters). Open an issue and discuss scope first — the book has one voice, and keeping it is deliberate.
  • Bulk AI-generated pull requests. If a tool drafted it and you didn't verify every line, don't submit it.
  • Style-only rewrites, reformatting, or "improvements" to prose that change no meaning.

How to contribute

  1. Small fixes (typos, a broken code line): open a pull request directly. One section per PR.
  2. Anything larger: open an issue first. Describe the problem before proposing the solution.

Ground rules for code in the book

  • NumPy-first. Framework code only where the section is explicitly about a framework.
  • Numerically stable: implementations must survive extreme inputs. Adversarial data is the whole point of this book.
  • Verified: if you change a function that has a derivative, the numerical gradient check must still pass. If you add one, add its check.
  • Every code block must run as-is when copied out of its YAML section, with only requirements.txt installed.

Format conventions

  • Book content lives in book/, one YAML file per section: section_NN_MM_description.yaml, indexed by chapter_NN_index.yaml.
  • Match the structure of neighboring sections: overview, concept blocks, security_implications, key_takeaways.
  • Every section connects to security. If a contribution doesn't answer "why does this matter for attack or defense," it isn't finished.

License

By contributing, you agree your contributions are licensed under the MIT License that covers the project.