📢 [Contribution Policy] Regarding AI-Generated Issues and Pull Requests

# Regarding AI-Generated Issues and Pull Requests

Thank you for your interest in the QwenPaw project and your enthusiasm for contributing. We truly value every member of this community.

Recently, we have noticed a growing number of issues and pull requests that appear to be entirely generated by AI tools (e.g., LLMs, coding agents) without meaningful human review. While we are open to — and even encourage — the use of AI-assisted development ("vibe coding"), **contributors bear the responsibility of being the first line of quality control**. Submitting fully automated, unreviewed content shifts the burden of validation entirely onto maintainers, which is unsustainable and counterproductive for everyone involved.

## Contribution Guidelines

Before submitting an issue or pull request, please ensure the following:

### For Issues

1. **Reproduce the problem yourself.** Confirm that the behavior you are reporting is a genuine, reproducible bug in QwenPaw — not a misconfiguration, a misunderstanding of intended behavior, or an AI hallucination.
2. **Provide your actual environment details.** Include your real QwenPaw version, OS, Python version, and steps to reproduce. Do not fabricate or let AI guess these details.
3. **Follow the issue template.** We provide issue templates for bug reports and feature requests. Please fill in every section — do not delete or skip fields. A well-formed issue is faster to triage and more likely to receive a timely response.

### For Pull Requests

1. **Verify the root cause.** Confirm that the issue you are fixing has been accurately diagnosed. Understand *why* the bug occurs, not just what the symptoms are.
2. **Validate the fix is correct and safe.** Ensure your changes do not break existing functionality. Beyond correctness, we encourage contributors to consider whether the fix aligns with the framework's overall architecture and represents a *minimal necessary change* — this leads to more elegant, maintainable solutions and is greatly appreciated during review.
3. **Test locally and show your work.** Run the relevant test suite and `pre-commit` checks. Include concrete evidence of your testing results (logs, screenshots, or command output) in the PR description.
4. **Follow the pull request template.** Our PR template includes sections for motivation, changes, and testing evidence. Please complete all sections. Incomplete or template-skipping PRs may be closed without review.

## On the Use of AI Tools

We welcome and encourage the use of AI tools to help analyze codebases, draft fixes, and accelerate development. However, this must be built on a foundation where **the contributor has carefully reviewed all AI-generated content** and can vouch for its correctness. Specifically:

- Verify that the AI's analysis is grounded in reality, not hallucinated code paths or imagined behaviors.
- Ensure the proposed solution actually solves the problem — test it, don't just trust it.
- Be prepared to explain and defend your changes in code review, as any human contributor would.

If your contribution is **entirely AI-generated and has not been verified by you**, please state this clearly in the issue or PR description. This allows maintainers to calibrate their review effort accordingly, and is far more respectful of everyone's time than submitting unreviewed content without disclosure.

## Why This Matters

Every issue and PR requires maintainer time to triage, review, and respond to. Unreviewed AI-generated contributions often contain subtle inaccuracies, misunderstand the codebase architecture, duplicate existing work, or propose changes that conflict with the project's direction. This doesn't just slow us down — it takes time away from meaningful contributions and genuine community discussions.

We are building QwenPaw together. Let's keep this community healthy, respectful, and productive.

Thank you for your understanding and cooperation.