# AI Usage Policy

> This policy is inspired by [Ghostty's AI Policy](https://github.com/ghostty-org/ghostty/blob/main/AI_POLICY.md).

We welcome the use of AI tools in this project. These guidelines exist to keep
contributions high-quality and to respect the time of human reviewers.

## Disclosure

All AI usage **must** be disclosed. In your pull request description, you must
include:

1. **Which tool(s)** you used (e.g., Claude Code, Copilot, Cursor, ChatGPT).
2. **What you used it for** — specify which parts of the code, docs, or tests
   were AI-assisted versus written by hand.
3. **How much** of the contribution is AI-generated (e.g., "AI wrote the initial
   draft, I rewrote the error handling and added tests").

Undisclosed AI usage that is discovered after the fact will be treated as a
policy violation.

## You Must Understand Your Changes

AI can help you write code, but you are responsible for every line you submit.
You must be able to explain what your changes do, why they are correct, and how
they interact with the rest of the codebase — without the AI tool. If you
cannot, the contribution is not ready. Maintainers may ask you to explain any
part of your submission, and inability to do so is grounds for closing the PR.

## Review AI-Generated Text

Issues, discussions, and documentation written with AI assistance must be
reviewed and edited by a human before posting. Remove boilerplate filler,
unnecessary verbosity, and irrelevant details. Keep it concise and useful.

## No AI-Generated Media

AI-generated images, art, audio, and video are not accepted. This policy covers
text and code only.

## Low-Effort Contributions

We do not accept low-effort contributions of any kind, AI-assisted or otherwise.
Examples include:

- Unreviewed AI output submitted as-is ("slop")
- Drive-by PRs that "fix" style or formatting without being asked
- Bulk changes with no clear purpose or explanation
- Issues or comments that are obviously AI-generated walls of text

Every contribution should demonstrate that the author has put genuine thought and
effort into it. If it looks like you spent more time prompting an AI than
understanding the problem, it will be closed.

## Enforcement

Any violation of this policy will result in the contribution being closed and
the contributor being **permanently blocked** from the project. There are no
warnings. A public record of blocked contributors may be maintained and shared
with other projects to help identify repeat offenders.

## Maintainers

These guidelines apply to external contributions. Project maintainers may use AI
tools at their own discretion.