What problem does it solve?
This Skill automates deep, repeatable codebase audits to discover security, dependency, code quality, architecture, performance, testing, and documentation issues while providing safe options for verification and optional auto-fixes.
Core Features & Use Cases
- Multi-mode audits: Supports parallel worktrees, single-session subagents, multi-clone, and manual orchestration for flexible isolation and scale.
- Findings compilation & issue creation: Aggregates agent results, de-duplicates findings, generates a human-readable audit report, and can create categorized GitHub issues linked to an epic.
- Safe auto-fix workflow: Optional --fix flow confines edits to git worktrees, runs verification (lint/type-check/build), commits per-agent branches, merges into a combined branch, and opens a PR upon success.
- Use Case: Run a thorough security and quality audit across a monorepo, then optionally apply high-confidence fixes and open a PR for review.
Quick Start
Run an interactive audit by invoking /audit to analyze the repository, review findings, and optionally create GitHub issues or an auto-fix PR.