review

Review code changes between branches and post findings to GitHub PRs.

9|1|Updated Feb 28, 2026
One-click install
npx skills add https://github.com/Roxabi/roxabi-plugins --skill review-roxabi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/Roxabi/roxabi-plugins/tree/main/plugins/dev-core/skills/review
Command: npx skills add https://github.com/Roxabi/roxabi-plugins --skill review-roxabi

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill automates the process of code review, providing detailed feedback on correctness, security, performance, and more, directly within the context of a pull request.

Core Features & Use Cases

  • Multi-Domain Analysis: Employs specialized agents (security, architecture, product, testing, etc.) to review code from various perspectives.
  • Conventional Comments: Generates findings in a structured format, including root cause, solutions, and confidence levels.
  • Automated Verdict: Provides a clear verdict (Approve, Approve with comments, Request changes) based on the findings.
  • PR Integration: Posts review comments directly to GitHub pull requests.
  • Use Case: After pushing a new feature branch, trigger the review skill to get immediate, comprehensive feedback before merging, ensuring code quality and adherence to standards.

Quick Start

Run the review skill on the current branch to analyze changes against the staging branch.

Frequently Asked Questions about review

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

FAQPage Schema
How do I automate code review on a GitHub pull request before merging?▼

To automate GitHub pull request code review, trigger the review skill on your current branch to analyze changes against staging. It then posts structured conventional comments and an automated verdict directly to your PR.

What is conventional comments format in automated code review?▼

Conventional comments in automated code review provide structured feedback by categorizing findings with metadata like root cause, proposed solutions, and confidence levels, ensuring clear and actionable pull request feedback.

Can I run a security and architecture analysis on my feature branch without manual checks?▼

Yes, you can run security and architecture analysis on a feature branch without manual checks by triggering automated review. Specialized agents evaluate multiple domains including correctness, performance, and observability simultaneously.

Does automated code review work with any GitHub branch or only pull requests?▼

Automated code review works by analyzing changes between any two branches, such as your current feature branch and staging. It then integrates directly with GitHub pull requests to post its findings and verdict.

What is the best way to get comprehensive testing and readability feedback on new code?▼

The best way to get comprehensive testing and readability feedback on new code is using an automated multi-domain review skill. It employs specialized agents to analyze correctness, tests, and readability, providing structured solutions.

How does an automated verdict decide whether to approve or request changes on a pull request?▼

An automated verdict decides whether to approve or request changes on a pull request by evaluating the structured conventional comments generated by specialized agents across security, architecture, and other code quality domains.