How does this skill differ from standard Claude coding?

It dispatches a specialized 'code-reviewer' subagent specifically trained to audit and critique code rather than just generate it, providing a more objective evaluation.

When is the best time to trigger a code review?

The skill recommends reviews after every task in subagent-driven development, after completing major features, or as a final check before merging to the main branch.

What are the feedback levels provided by the reviewer?

Issues are categorized as Critical (fix immediately), Important (fix before proceeding), or Minor (note for later improvement).

Can I push back if the reviewer subagent is wrong?

Yes, if the reviewer provides incorrect feedback, you are encouraged to push back with technical reasoning or code evidence to clarify the implementation.

What technical information is required to start a review?

You need to provide the BASE_SHA and HEAD_SHA of the git commits being reviewed, along with a description of the implementation and the original requirements.

Requesting Code Review

Name: Requesting Code Review
Author: yang0

byyang0

0•

协作工具

Dispatches a specialized code-reviewer subagent to verify code quality, identify bugs, and ensure implementations meet project requirements.

This skill streamlines the quality assurance process by allowing Claude to invoke a dedicated 'code-reviewer' subagent at critical development milestones. By leveraging Git SHAs and structured templates, the skill provides the reviewer with the necessary context to compare new implementations against original plans and requirements. It helps prevent technical debt and cascading errors by mandating reviews after individual tasks or major features, providing actionable feedback categorized by severity to ensure only high-quality code reaches the main branch.

主要功能

010 GitHub stars

02Categorized feedback labels (Critical, Important, Minor) for prioritization

03Contextual reviews using Git SHA diffs and requirement documents

04Integration with subagent-driven development for incremental quality checks

05Automated subagent dispatch for objective, specialized code analysis

06Template-based review requests to ensure consistent evaluation criteria

使用场景

01Verifying major feature implementations before merging into the production branch

02Gaining a fresh perspective on complex bug fixes or refactoring efforts

03Performing incremental 'smoke tests' after completing individual tasks in a development plan

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yang0/superpower requesting-code-review

For use in Claude.ai and ChatGPT

主要功能

010 GitHub stars

02Categorized feedback labels (Critical, Important, Minor) for prioritization

03Contextual reviews using Git SHA diffs and requirement documents

04Integration with subagent-driven development for incremental quality checks

05Automated subagent dispatch for objective, specialized code analysis

06Template-based review requests to ensure consistent evaluation criteria

使用场景

01Verifying major feature implementations before merging into the production branch

02Gaining a fresh perspective on complex bug fixes or refactoring efforts

03Performing incremental 'smoke tests' after completing individual tasks in a development plan

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add yang0/superpower requesting-code-review

For use in Claude.ai and ChatGPT