We welcome contributions that used AI tools, provided they meet our quality, testing, security and licensing expectations. AI-generated code is treated the same as human-written code — it must be reviewed, tested, and owned by the contributor.
If you used AI to produce any part of your contribution, include the following in your PR description:
- Model/Agent (e.g., "GPT-5 Thinking mini via Copilot")
- How AI was used (e.g., "boilerplate, implementation draft, tests, docs")
- Manual validation performed (linters, tests, security checks)
Simple autocompletion does not require disclosure. Substantial generation (functions, algorithms, large refactors, tests, docs) requires disclosure.
- You remain fully responsible for the code you submit. Understand, test, and be able to explain all changes.
- All code (AI or human) must pass linters and tests and meet project coding standards.
- Provide documentation and tests for non-trivial changes.
- By contributing you confirm you have the right to contribute the content (including AI outputs) under this project's license.
- Do not submit content that includes proprietary, copyrighted, or secret data.
- Avoid insecure patterns and secrets in contributions.
- For core, API, or architectural changes open an Issue first and discuss; link the Issue from the PR.
- Small bug fixes and docs may open PRs directly, but still follow disclosure and testing requirements.
Maintainers may reject or modify PRs that diverge from project goals, introduce undue complexity, or violate the above rules.
This policy may be updated as the ecosystem evolves. If maintainers suspect problematic AI-generated content they may request provenance, tests, or rework.