ac-review

Automate code review for repository changes and write markdown reports.

6|2|Updated Jan 9, 2026
One-click install
npx skills add https://github.com/yz0812/ai-dev-toolkit --skill ac-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ac-review
Source: https://github.com/yz0812/ai-dev-toolkit/tree/main/skills/ac-review
Command: npx skills add https://github.com/yz0812/ai-dev-toolkit --skill ac-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates code review tasks within a Claude Code workflow, generating structured feedback and actionable plans without manual review overhead.

Core Features & Use Cases

  • Automatic git-diff review: inspect changes and surface issues, potential regressions, and improvement suggestions.
  • Plan generation after scope confirmation: write both a review report and a remediation plan to .claude/review and .claude/plan.
  • Role separation and workflow orchestration: main thread coordinates subagents to fetch context, run checks, and aggregate results without modifying source code.

Quick Start

Run /ac-review to audit the current git changes and generate a remediation plan.

Frequently Asked Questions about ac-review

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

FAQPage Schema
How do I automate code review for git diffs in a Claude Code workflow?

You can automate code review for git diffs by running a command that inspects repository changes, surfaces regressions, and writes structured markdown reports to a designated directory without modifying source code.

Can I generate a remediation plan automatically after a code review?

Yes, after the initial review confirms scope, a separate remediation plan is generated and written as a markdown file to the project directory, providing actionable steps to fix identified issues.

Does automated code review with subagents modify my source code?

Automated code review with subagents maintains a read-only orchestration process, meaning it fetches context and aggregates results without modifying your actual source code files.

What is the best way to review specific git revisions and generate markdown reports?

The best way to review specific revisions is using a frontmatter-driven workflow that audits the changes, identifies potential improvements, and outputs both review reports and remediation plans in markdown format.

How does subagent orchestration work for automated code review?

Subagent orchestration for code review works by having the main thread coordinate subagents to fetch context and run checks, aggregating results into structured markdown outputs without touching source code.

Do I need to manually trigger code review and plan generation separately?

No, you do not need to manually trigger them separately; a single command audits current git changes and generates both the review report and the remediation plan sequentially after scope confirmation.