What problem does it solve?
This Skill prevents risky or low-quality code from being committed by running a pre-commit verification pipeline that combines security scanning, baseline-aware checks, and an independent reviewer loop.
Core Features & Use Cases
-
Static security scan on the diff: Flags likely hardcoded secrets and common injection patterns from added lines before anything is committed.
-
Baseline-aware quality gates: Compares test and lint outcomes against the state before your changes to avoid failing on pre-existing issues.
-
Independent reviewer subagent + auto-fix loop: Sends only the diff and scan results to a fresh reviewer for fail-closed verdicts, then optionally runs up to two targeted fix-and-reverify cycles.
-
Use Case: After changing multiple files in a git repo (for example, implementing a feature that touches both backend logic and tests), run this to catch security regressions and ensure new failures are addressed before pushing or opening a PR.
Quick Start
Run the code-review verification after making changes and before committing by telling the AI to verify your staged diff for a safe, baseline-aware quality gate.
