review

Review git diffs with multi-agent detection of correctness, security, performance, and style issues.

26|8|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/arbazkhan971/godmode --skill review-arbazkhan971
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: review
Source: https://github.com/arbazkhan971/godmode/tree/main/skills/review
Command: npx skills add https://github.com/arbazkhan971/godmode --skill review-arbazkhan971

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides a fast, consistent, and actionable code review process that finds correctness, security, performance, and style issues across a branch diff, reducing manual review time and ensuring issues are recorded with concrete fixes.

Core Features & Use Cases

  • Multi-agent parallel review: Four specialized agents run concurrently to detect correctness, security, performance, and style problems.
  • Deduplication and severity ranking: Findings from agents are merged, deduplicated, and sorted by severity with clear file:line evidence and suggested code fixes.
  • Safe auto-fixes and audit trail: NIT-level fixes are applied in separate commits, tests guard auto-fixes, and logs are appended to .godmode/review-log.tsv for traceability.
  • Use Case: Run a review on a feature branch before merge to auto-fix trivial issues, block merges on MUST-FIXs, and generate a prioritized remediation list for the author.

Quick Start

Run /godmode:review to review the current branch diff with four agents, apply safe autofixes for NITs, and output a deduplicated, severity-ranked findings list.

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 for git pull requests?

Automate code review for git pull requests by running a multi-agent workflow that analyzes branch diffs for correctness, security, performance, and style issues, outputting file:line findings with suggested fixes and severity rankings.

How does multi-agent code review work for branch comparisons?

Multi-agent code review works by running four specialized agents concurrently across your branch diff to detect correctness, security, performance, and style problems. Results are merged, deduplicated, and severity-ranked with clear evidence.

Can I automatically fix trivial code issues found in a git diff?

Yes, you can automatically fix trivial code issues by applying NIT-level autofixes in separate commits. Tests guard these auto-fixes to ensure safety, and an audit trail is logged for traceability.

How do I handle large git diffs during automated code review?

Handle large git diffs during automated code review by splitting the diff per-directory before processing. This allows the parallel agents to analyze large codebases efficiently without missing issues.

Does automated code review stop when critical security or correctness issues are found?

Yes, automated code review enforces stop conditions when failures occur. MUST-FIX severity issues block merges, ensuring critical correctness and security problems are addressed before proceeding.