Cavet Triage

Interpret security scan findings, separate false positives from real issues, and surface only what matters. Use after code is written or changed, before a commit, whenever scan output appears in context (including pre-commit hook output), when the operator asks to check for vulnerabilities, "is this safe", "run a scan", or asks what to do about a finding. Covers both the isolated triage subagent and the parent session that reconciles its results.

ChaosChild Updated
cavet staged scan finds a planted key, the finding is dismissed with a recorded reason, and the audit trail shows every event

File contents

ChaosChild/cavet commit 2a05464109

Frequently asked questions

npx get-skill add chaoschild/cavet-triage