How does the Audit skill handle large codebases?

The skill manages high token usage by prompting the user to run the /compact command first and utilizes background tasks to process multiple domains simultaneously without blocking the main thread.

Can I run the audit on just a specific part of my project?

Yes, the skill supports scoped audits. You can provide arguments like 'audit auth', 'audit dashboard', or 'audit latest' to focus the agents on specific features or recent changes.

What specific areas are checked during the audit?

The audit covers seven domains: Security (secrets, injections), Performance (memoization, re-renders), Accessibility (WCAG compliance), Type Safety (any usage, ts-ignore), UX/UI (loading states, responsive design), Test Coverage, and Deploy Readiness.

Does the skill automatically fix the issues it finds?

It generates a detailed report and creates actionable stories in your prd.json. You can then use the 'auto' command or other development skills to address the findings based on their assigned priority.

Parallel Quality & Security Audit

Name: Parallel Quality & Security Audit
Author: djnsty23

bydjnsty23

•

セキュリティとテスト

Executes a multi-agent quality audit across seven domains to ensure production readiness and identify critical vulnerabilities.

The Audit skill orchestrates a swarm of seven specialized Claude Opus agents to perform a comprehensive deep-dive into your codebase. It simultaneously analyzes security, performance, accessibility, type safety, UX/UI consistency, test coverage, and deployment readiness. By aggregating findings into a detailed report and automatically persisting actionable stories to your project management files, it transforms raw analysis into a structured development roadmap. This skill is ideal for pre-release checks or when taking over a new repository to quickly identify technical debt and high-priority fixes.

主な機能

01Automated task creation and persistence to prd.json for project tracking

02Parallel execution of 7 specialized Opus agents for maximum efficiency

03Context-aware analysis with support for full or scoped audits

04Deep security scanning for secrets, XSS, and Supabase RLS policies

05Comprehensive accessibility (WCAG) and UX/UI feedback loops

062 GitHub stars

ユースケース

01Identifying security vulnerabilities and hardcoded secrets in existing code

02Final quality gate before major production releases

03Mapping technical debt and missing test coverage in new or inherited projects

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add djnsty23/claude-auto-dev audit

For use in Claude.ai and ChatGPT

Download Skill