How does the skill handle large code changes?

For diffs exceeding 200 lines, the skill automatically activates 'Thorough Mode,' which spawns three specialized sub-agents to analyze specifications, clean code, and overengineering in parallel.

Can I review my implementation against a specific design doc?

Yes, by using the --spec flag, you can provide a reference document that the skill will use to validate that your code does exactly what was requested.

Where are the code review reports saved?

When the --report flag is used, findings are saved as formatted Markdown files in the docs/reviews/ directory, organized by date.

What is the difference between Single-Pass and Subagent mode?

Single-Pass mode is a quick, cost-effective review for small changes. Subagent mode uses specialized agents to perform deeper analysis on complex logic and architectural patterns.

Does it support branch comparisons?

Yes, using the --branch flag, the skill can calculate the merge-base with the main branch and review all changes made on your current feature branch.

Code Quality Reviewer

Name: Code Quality Reviewer
Author: rickym-h

byrickym-h

0•

协作工具

Performs automated, tiered code reviews to ensure specification adherence, clean code principles, and architectural integrity.

The Code Quality Reviewer skill automates the rigorous process of code analysis within Claude Code, ensuring every change aligns with project specifications and best practices. It intelligently selects between a fast single-pass review for small updates and a powerful multi-agent 'Thorough Mode' for large diffs, deploying specialized agents to hunt for overengineering, YAGNI violations, and maintenance bottlenecks. By providing actionable, file-specific feedback categorized by severity, it empowers developers to catch critical issues early and maintain high-quality codebases with minimal manual overhead.

主要功能

01Intelligent dual-mode analysis based on diff size or user preference

020 GitHub stars

03Multi-agent parallel review for thorough architectural assessments

04Strict requirement validation against provided specification documents

05Automated detection of overengineering and premature optimization

06Tiered reporting (Critical, Important, Minor) with file-line references

使用场景

01Verifying that new feature implementations strictly adhere to technical specification documents

02Automating pre-pull request checks to identify clean code violations before human review

03Auditing large architectural refactors for maintainability and separation of concerns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rickym-h/jadefall-marketplace review

For use in Claude.ai and ChatGPT

Download Skill