copilot-review

Route git diffs through Copilot sub-agents for multi-perspective code reviews.

Updated Mar 17, 2026
One-click install
npx skills add https://github.com/g-kari/g-kari-plugins --skill copilot-review-g-kari
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: copilot-review
Source: https://github.com/g-kari/g-kari-plugins/tree/main/copilot-review/skills/copilot-review
Command: npx skills add https://github.com/g-kari/g-kari-plugins --skill copilot-review-g-kari

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill enables fast, thorough code reviews by orchestrating Copilot CLI across multiple review perspectives in parallel, reducing manual review time and surfacing issues early.

Core Features & Use Cases

  • Multiview review: simultaneously run five perspectives (bug/logic, security, error handling, performance, readability/maintainability) on a given diff.
  • Source-agnostic targets: works with the latest changes, a specific commit range, or a PR diff, across unstaged or staged changes.
  • Consolidated feedback: aggregates results from all perspectives into a unified report with actionable fix suggestions.

Quick Start

Feed the diff to Copilot and request a parallel, multi-perspective review.

Frequently Asked Questions about copilot-review

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

FAQPage Schema
How do I run a multi-perspective code review on a git diff?

To run a multi-perspective code review, you feed the git diff to Copilot CLI, which routes the changes through parallel sub-agents to evaluate bug/logic, security, error handling, performance, and readability simultaneously. It aggregates the results into a consolidated report with actionable fix suggestions.

Can I review staged and unstaged changes across specific commit ranges?

Yes, you can review staged and unstaged changes across specific commit ranges. The skill applies source-agnostic targeting to analyze the latest diffs, specific commit hashes, or PR diffs by processing the git diff output through parallel Copilot agents.

What is the best way to automate security and performance checks on a PR diff?

The best way to automate security and performance checks is routing the PR diff through Copilot CLI with parallel sub-agents. This approach simultaneously evaluates multiple review perspectives like security and performance, consolidating the feedback into a unified report.

Does this parallel review approach work with GitHub Copilot CLI?

Yes, this parallel review approach works directly with GitHub Copilot CLI by integrating with the gh copilot -p command. It orchestrates multiple AI agents to collect and present a consolidated per-perspective report from the Copilot output.

How do I consolidate feedback from different code review perspectives into one report?

You consolidate feedback by routing diffs through Copilot with parallel sub-agents that handle bug/logic, security, error handling, performance, and readability. The system automatically aggregates results from all perspectives into a unified report with actionable fix suggestions.