What is the Requesting Code Review skill?

It is a specialized Claude Code capability that dispatches a subagent to review your code changes specifically against your project requirements or development plans.

How does the skill access my code changes?

The skill utilizes Git SHAs to identify the specific range of commits you wish to have reviewed, allowing the subagent to see exactly what was implemented.

When is the best time to request a review?

It is mandatory after each task in subagent-driven development and before merging to main, but also valuable when stuck or before starting a major refactor.

How do I act on the feedback provided by the subagent?

Feedback is categorized by severity. You should fix Critical and Important issues immediately, while Minor issues can be noted for later or addressed if time permits.

Can I disagree with the AI code reviewer?

Yes. If the reviewer is incorrect, you can push back with technical reasoning, provide tests, or ask for clarification to refine the review process.

AI Code Review Orchestrator

Name: AI Code Review Orchestrator
Author: adawalli

byadawalli

0•

协作工具

Initiates a specialized subagent to audit code implementations against requirements and catch issues before they merge.

This skill streamlines the quality assurance process within Claude Code by dispatching a dedicated code-reviewer subagent to evaluate work. By comparing git diffs against original project plans or requirements, it ensures that implementations are not only functional but also aligned with the intended scope. It provides a structured framework for identifying critical bugs, architectural flaws, and minor improvements early in the development lifecycle, preventing technical debt from accumulating during subagent-driven or ad-hoc development tasks.

主要功能

01Automated subagent dispatch for deep code auditing

02Seamless integration with subagent-driven development workflows

030 GitHub stars

04Verification of implementation against specific project requirements

05Git-aware analysis using HEAD and Base commit SHAs

06Categorized feedback loop for Critical, Important, and Minor issues

使用场景

01Verifying task completion in multi-step development plans

02Performing a final audit of major features before merging to the main branch

03Getting a fresh perspective on complex bug fixes or architectural refactors

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add adawalli/claude-plugins requesting-code-review

For use in Claude.ai and ChatGPT

Download Skill