Thank you for your interest in PicoLimbo!
I am currently the primary maintainer of PicoLimbo. To keep development efficient, I ask that you do not submit pull requests generated by LLMs.
I use AI myself for autocomplete and small tasks, but "vibe coding" (submitting large blocks of AI-generated code without deep understanding) creates an unsustainable amount of review work for a solo maintainer.
If you use AI to assist you, please ensure:
- You have personally verified and understood every line of the change.
- You must disclose how the AI was used in your PR description.
- You are prepared to discuss the implementation details during review.
Please try to keep pull requests as focused as possible. A PR should do exactly one thing. If a PR touches unrelated features or refactors unrelated code, it will likely be rejected.
Smaller PRs are easier to review, faster to merge, and less likely to introduce regressions.
PicoLimbo aims to support a vast range of versions. Because of this, comprehensive testing is mandatory.
Before submitting a PR, you must verify two things:
- Feature verification: Your new feature or fix works exactly as intended.
- Regression testing: Your changes have not broken existing functionality in other versions.
In your PR description, you must explicitly state which versions you have tested.
- Example: "Tested on 1.21 and 1.20. This feature is not compatible with versions below 1.20, and I have not tested those."
- If you cannot test a specific version range, you must state that clearly.