Where are the audit results saved?

The skill generates a detailed markdown report and saves it to the docs/audits/ directory, which is then automatically committed to your git history.

Can I audit a specific part of my repository?

Yes, the audit skill accepts a path scope, allowing you to focus the review on specific directories like 'apps/web' or 'src/components'.

Does it check for known security vulnerabilities?

Yes, it includes a security sentinel agent and automatically runs dependency scans (npm/pip) to surface critical and high-severity vulnerabilities.

How does it handle large codebases without hitting limits?

By default, the skill runs reviewers in batches of two to prevent resource exhaustion, though you can use the --parallel flag to run them all at once if your system allows.

What frameworks does the audit skill support?

It features specialized logic for Next.js, React, TypeScript, Python, Rust, and Go, and can detect databases like Prisma or Drizzle for data integrity checks.

Comprehensive Codebase Audit

Name: Comprehensive Codebase Audit
Author: howells

byhowells

•

Security & Testing

Performs multi-dimensional codebase reviews using specialized AI agents to identify security, performance, and architectural issues.

The Audit skill automates the complex process of code review by deploying a fleet of specialized AI 'reviewer' agents tailored to your project's specific scale and technology stack. It intelligently detects your framework—whether it's Next.js, React, Python, or Go—and applies domain-specific logic to find security vulnerabilities, performance bottlenecks, and architectural debt. By consolidating findings from multiple experts into a prioritized markdown report, it provides a clear roadmap for improving code quality and system reliability.

Key Features

016 GitHub stars

02Comprehensive Markdown report generation with prioritized action items

03Specialized reviewer agents for Security, Performance, Architecture, and UI/UX

04Integration with .ruler files for project-specific coding standards enforcement

05Dependency vulnerability scanning for Node.js and Python projects

06Automated project scale detection to adjust audit depth and resource usage

Use Cases

01Pre-production security and performance clearance for critical releases

02Technical debt assessment when onboarding onto a legacy or unfamiliar codebase

03Identifying and refactoring 'AI slop' or low-quality patterns in AI-assisted development

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add howells/arc audit

For use in Claude.ai and ChatGPT

Download Skill