skill-authoring
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of a single documentation file providing structural guidelines, templates, and conventions for authoring other agent skills.
- [SAFE]: It contains instructions on metadata optimization (CSO rule), size budgeting, and safety-oriented patterns like 'Iron Laws' and 'Red Flags' designed to improve agent reliability.
- [SAFE]: References to internal scripts and files (e.g., '.github/scripts/validate_skills.py', '.github/skill-budgets.json', and 'tests/python/test_plugin_manifest.py') are for repository maintenance and continuous integration validation, not for execution of untrusted code.
Audit Metadata