Purpose and inputs
Find security defects relative to _first/.../SECURITY.md, repository security docs, and existing scanners. Do not invent thresholds. Stay report-only unless the user asked to fix.
Steps
- Load
/f-securityand the Security overlay. If missing, stop and ask; do not invent a bar. - Inspect dependencies, authz, secrets handling, and input boundaries that this repo already names.
- Run existing security scripts or CI jobs from the overlay or package.json. Record passed, failed, or not run.
- If fixes are authorized, change the owning cause. Policy changes go through
/f-security.
Verification
- Each finding cites overlay, check, or source evidence.
- No new password, encryption, or header policy was introduced.
- No commit unless the user asked.
Handoff
List findings, residual risk, and whether /f-security must own a follow-up.