review

Reviews code changes against correctness, security, performance, and style checklists to generate structured findings reports.

4|2|Updated Jan 27, 2026
One-click install
npx skills add https://github.com/Arete-Consortium/ai-skills --skill review-arete-consortium
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/Arete-Consortium/ai-skills/tree/main/personas/engineering/review
Command: npx skills add https://github.com/Arete-Consortium/ai-skills --skill review-arete-consortium

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Code reviews are often inconsistent, with reviewers missing security vulnerabilities, edge cases, or test gaps. This Skill provides a systematic checklist-driven review process that produces structured findings with severity levels and line references. ## Core Features & Use Cases - Comprehensive Checklists: Covers correctness, code quality, testing, security, performance, documentation, and style in a single review pass. - Focused Review Modes: Run targeted security, performance, or API reviews with dedicated checklists for each domain. - Structured Findings Report: Generates findings categorized as Critical, Major, and Minor with file locations and concrete fix suggestions, plus a verdict (Approve, Request Changes, or Comment). - Use Case: Before merging a pull request, invoke the review to scan the diff for SQL injection risks, missing error handling, and untested edge cases, then receive a formatted report with blocking issues and constructive feedback. ## Quick Start Ask Claude to review the staged changes or a specific file using the review checklist and report any security or correctness issues.

Frequently Asked Questions about review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I review a pull request with Claude Code?▼

Invoke the review command with the PR number, such as /review --pr 123, or run /review on staged changes. Claude analyzes the diff, checks correctness, security, and tests, then generates a findings report with severity levels and a verdict.

What does a code review checklist cover?▼

The checklist covers correctness, code quality, testing, security, performance, documentation, and style. Each category contains specific items like input validation, N+1 query detection, test coverage, and naming consistency.

Can I run a security-only code review?▼

Yes, use the --security flag to focus the review on vulnerabilities. The security checklist checks input validation, SQL injection, XSS, hardcoded secrets, authentication, authorization, and CSRF protection.

How are code review findings reported?▼

Findings are grouped into Critical (must fix), Major (should fix), and Minor (nice to have) tables with file locations, issue descriptions, and suggested fixes. The report ends with a verdict: Approve, Request Changes, or Comment.

What are the limitations of checklist-based code review?▼

Checklist reviews catch common patterns but may miss domain-specific logic errors or architectural issues requiring deep system context. They complement rather than replace human judgment on design decisions and business logic correctness.