Skip to content

kube-burner logo

kube-burner LLM (AI) Contribution Policy

Large language models (LLMs) such as ChatGPT, Copilot, and Claude can help contributors work faster. They can also introduce risk when output is submitted without careful review.

Kube-burner values clear communication, maintainable code, and high review quality. This policy defines expectations when AI tools are used in kube-burner repositories and community spaces.

Direct Communication

Do not post raw or copy-pasted LLM output as your own communication.

This applies to:

  • GitHub issues and issue comments
  • Pull request descriptions and review comments
  • Commit messages
  • Slack, discussions, and other project communication channels
  • Security and vulnerability reports

Write in your own words and make sure you understand what you submit.

Exceptions

  • Translation help is allowed. If you used an LLM for translation, say so clearly (for example: "Translated with an LLM from ").
  • Maintainer-managed automation or bots may suggest changes. These suggestions are not authoritative and must still be reviewed by humans.

Repeated violations may result in closed submissions or removal from project spaces.

Code Contributions Assisted by LLMs

Using LLMs for coding is allowed, but responsibility remains with the contributor.

Requirements

  • Follow all project contribution guidance in CONTRIBUTING.md
  • Keep changes focused, minimal, and easy to review
  • Match existing style, formatting, and structure
  • Remove unnecessary noise (unused files, generated metadata, unrelated edits)
  • Ensure your change builds and passes relevant tests before requesting review
  • Test the behavior you changed

Ownership and Understanding

You must:

  • Review generated code before submitting
  • Explain what changed and why in your own words
  • Be able to discuss implementation details during review

Submissions that are clearly not understood by the author may be rejected.

Responding to Review Feedback

Do not blindly paste review comments into an LLM and submit the response unchanged.

Contributors are expected to:

  • Respond thoughtfully in their own words
  • Make targeted updates that address the feedback
  • Understand and validate each requested change

Maintainer Discretion

Maintainers may decline changes that are difficult to review, overly large, poorly structured, or repeatedly low quality, regardless of whether AI tools were used.

Golden Rule

Use LLMs as assistants, not as substitutes for understanding, accountability, and craftsmanship.