Skip to content

Latest commit

 

History

History
34 lines (23 loc) · 2.17 KB

File metadata and controls

34 lines (23 loc) · 2.17 KB

Contribution guidelines:

General guidelines

Keep everything respectful; conflicts don't move the project forward.
Listen to requests and recommendations from the maintainers.
Make sure a duplicate of your pull request doesn't already exist.
Do not reopen closed pull requests marked as not planned.

If you contribute code to this project, you agree that this code is now licenced under the GPL 3.0.

AI Policy

AI-generated and agentic PRs

Not allowed: Fully automated or agentic PRs (e.g. autonomous coding agents like Claude Code submitting a PR on your behalf). These will be closed without review, and repeated attempts may result in a ban.

Allowed: AI-assisted coding: autocomplete, or supervised prompting where you generate small chunks of code and review each one yourself.

If you use AI assistance, you must still:

  • Ensure the code meets this codebase's quality standards.
  • Take full responsibility for everything the AI produced. "the AI wrote it" is not a defense for bugs, bad style, or security issues.

We know AI tools are genuinely useful for coding. These limits exist to prevent abuse, spaghetti code, and PR volume that humans can't realistically review.

AI-generated and agentic issues

Not allowed: Issues that are fully AI generated or filed by autonomous/agentic tools without a human actually reproducing and verifying the problem first. This includes reports auto-filed from CI bots, LLM-driven "scanning" agents, or copy-pasted model output with no human review. Allowed: Using AI to help you draft, translate, or clean up an issue you've personally experienced and verified.

If you use AI assistance to write an issue, you must still:

  • Confirm the bug is real and reproducible on your end before filing.
  • Include actual steps to reproduce, logs, and environment details, not just a model's guess at what might be wrong.
  • Take full responsibility for the accuracy of the report. Issues that turn out to be hallucinated or unreproducible will be closed, and repeated low-effort or bot-filed issues may result in a ban.

We expect issue volume like this to increase, so this policy will be enforced the same way as the AI code policy above.