How does the review-all skill handle different programming languages?

The skill utilizes specialized review commands that interpret codebase context dynamically, making it language-agnostic while applying specific best practices for backend, frontend, and infrastructure.

Does it automatically fix the issues it discovers?

It identifies issues and creates detailed todo files with proposed solutions and acceptance criteria, allowing the developer to review, triage, and implement fixes.

What happens to low-priority findings?

Findings marked as 'NIT' are included in the final report for visibility but are skipped during the automatic todo creation process to prevent backlog clutter.

Can I see a summary of the findings before they become tasks?

Yes, the skill produces a unified assessment report grouped by category and sorted by severity before appending the count of todos created in your .claude/todos/ directory.

How long does a full 30-point review take?

Because the 30 commands run in parallel using Task agents, the process is highly optimized and typically completes in a few minutes, even for significant code changes.

Comprehensive Code Reviewer

Name: Comprehensive Code Reviewer
Author: jayteealao

byjayteealao

0•

보안 및 테스팅

Performs an exhaustive 30-point automated code review covering security, architecture, performance, and UX in parallel.

The review-all skill is a powerhouse for quality assurance, orchestrating 30 specialized review agents to analyze your codebase simultaneously. It evaluates everything from logical correctness and security vulnerabilities to infrastructure configuration and user experience. After processing, it merges findings into a prioritized report and automatically populates a local todo system within your directory, allowing developers to systematically address blockers and high-priority issues before deployment. It is the ultimate tool for ensuring production-readiness and maintaining high engineering standards across complex projects.

주요 기능

01Final assessment logic to determine 'Ship' or 'Don't Ship' status

02Unified report merging with severity-based prioritization (Blocker to Nit)

03Automatic generation of actionable .claude/todos files with metadata

04Parallel execution of 30 distinct review modules for maximum speed

05Comprehensive coverage of Security, DevOps, Architecture, and UX domains

060 GitHub stars

사용 사례

01Auditing legacy codebases for hidden security vulnerabilities or architectural drift

02Automating the population of technical debt backlogs during development sprints

03Performing a final safety audit before merging critical Pull Requests

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jayteealao/agent-skills review-all

For use in Claude.ai and ChatGPT

Download Skill