This skill helps AI agents follow best practices to confirm before executing, publish only with approval, and avoid over-automation for reliable results.
- Trust
- Needs review
- Install
- MCP ready
- Top risk
- Medium risk · Important install, trust, or metadata evidence is still thin.