Thank you for your interest in contributing! This project welcomes improvements from the community.
- Use the GitHub issues
- Include your AI client (Claude, Cursor, Kiro, etc.), OS, and Python version
- Provide a minimal reproduction when reporting bugs
- Fork the repository
- Create a feature branch:
git checkout -b feature/my-improvement - Make your changes
- Test
scripts/arc.pyparses:python3 -c "import ast, sys; ast.parse(open('scripts/arc.py').read())" - Commit with a clear message
- Open a PR against the main branch
| Area | Ideas |
|---|---|
| Skills | New agent-specific skills (Section 1 changes only) |
| Engine | Bug fixes, performance improvements, new CLI commands |
| Integrations | Rules files for additional AI tools |
| Docs | README improvements, new examples, clarity edits |
| Security | New scan rules, compliance mappings |
- Universal content (Sections 2–10 of any SKILL.md): Make the change in one SKILL.md first, then replicate to the other two.
- Agent-specific content (Section 1): Modify only the relevant agent's file.
- Engine changes (
scripts/arc.py): All skills reference the same script — update once.
- Python: follow PEP 8; the engine is stdlib-only — no external dependencies
- Markdown: 80-character soft wrap for readability
- Skills: keep each SKILL.md self-contained; AI agents consume them as single documents