Code Mower is built around careful technical disagreement: different AI systems, reviewers, and humans should be able to challenge one another without making the workspace hostile.
- Be respectful and specific when giving feedback.
- Assume good intent, but correct unsafe behavior quickly.
- Keep discussions focused on evidence, reproducible commands, repository context, and user impact.
- Redact API tokens, private source, raw diffs, raw model transcripts, credentials, auth output, and customer data before posting public examples; use SECURITY.md for private reports.
- Accept that maintainer decisions may prioritize safety, privacy, cost control, or first-user reliability over a clever integration.
- Harassment, threats, discrimination, or personal attacks.
- Publishing someone else's private code, credentials, identity details, or proprietary repository information.
- Pressuring maintainers or users to run untrusted commands, bypass security controls, or expose provider credentials.
- Using issues, discussions, or PRs to launder prompt injection, secret exfiltration, or other unsafe automation patterns.
For conduct concerns, contact the maintainers privately. For security issues, use the process in SECURITY.md instead of opening a public issue.
Maintainers may remove content, lock threads, block accounts, or revoke access when needed to protect users and the project.