# Secure AI Agent Coding

This wrapper exists for repository presentation. The authoritative instructions live in `SKILL.md`.

Read `SKILL.md` first, then load only the file that matches the task:

- `references/review-workflow.md` for reviews, incidents, and report shape
- `references/controls.md` for control evidence and checklist work
- `references/implementation-patterns.md` for implementation guidance
- `references/threat-model.md` for AI-specific threat modeling
- `references/governance.md` for inventory, lifecycle, data, consent, and operations
- `references/gotchas.md` for failure modes and reviewer traps
- `references/source-policy.md` for source conversion notes