This project recognizes that AI-assisted coding is a powerful tool for modern development. To ensure the quality, security, and maintainability of our codebase, we have established the following guidelines regarding the use of Artificial Intelligence (LLMs, Copilots, etc.).
You are not prohibited from using AI tools to help write code, generate documentation, or brainstorm solutions. We care about the final result: high-quality, functional, and secure code.
All contributions must be "vouched for" by a human.
- Human Responsibility: Every commit and Pull Request (PR) must have a primary human author/committer who fully understands every line of code being submitted.
- Vouching: By submitting a PR, you are stating that you have reviewed the AI output, verified its correctness, and take full responsibility for its inclusion in the project.
If a commit or PR is heavily generated or assisted by an AI, we recommend (but do not strictly require) acknowledging the tool in the commit metadata.
- You can add a trailer to your commit message like so:
Co-authored-by: AI-Assistant <ai-bot@example.com> - Regardless of attribution, the Main Committer must always be a human.
- No AI-Only Reviews: Pull Requests cannot be approved or merged based solely on the output of AI-based review tools.
- Human Oversight: A human maintainer or contributor must perform a manual review and "vouch" for the PR before it is accepted.
Not necessarily. Because our policy requires a human to be the primary committer and "vouchee" for every single line of code, the origin of the logic is secondary to the human's verification of it. As long as you understand and stand behind the code as if you wrote it yourself, the human verification is perfectly sufficient.
Ensuring our code remains readable, maintainable, and human-verified. 🚀