bmad-code-review

Review code changes with parallel review layers and structured triage.

Updated Mar 30, 2026
One-click install
npx skills add https://github.com/GenEducation/GenedUIProject --skill bmad-code-review-geneducation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-code-review
Source: https://github.com/GenEducation/GenedUIProject/tree/main/.gemini/skills/bmad-code-review
Command: npx skills add https://github.com/GenEducation/GenedUIProject --skill bmad-code-review-geneducation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams review code changes more thoroughly by running an adversarial, multi-layer review process that surfaces bugs, edge cases, and spec mismatches before merge.

Core Features & Use Cases

  • Parallel review layers: Compares findings from a blind reviewer, an edge-case hunter, and an acceptance auditor to catch different classes of defects.
  • Structured triage: Deduplicates, classifies, and prioritizes findings into patch, defer, dismiss, or decision-needed outcomes.
  • Workflow coordination: Handles diff selection, spec loading, story-file updates, deferred work tracking, and sprint-status syncing.
  • Use case: A developer asks for a review of a branch or commit range and gets an actionable report that separates real fixes from noise.

Quick Start

Use the bmad-code-review skill to review the selected code changes against the relevant spec and summarize actionable findings.

Frequently Asked Questions about bmad-code-review

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

FAQPage Schema
What is adversarial code review and how does it find edge cases in pull requests?

Adversarial code review uses parallel review layers like a blind reviewer, an edge-case hunter, and an acceptance auditor to catch different defect classes in pull requests before merge.

How do I review uncommitted diffs against acceptance criteria before shipping?

You review uncommitted diffs by loading the relevant spec and running multi-role analysis that compares code changes against acceptance criteria to surface spec mismatches and bugs.

Can I use this code review process for staged changes and branch commits?

Yes, the code review process applies to staged changes, uncommitted diffs, branch commits, and pull requests, parsing diffs to perform multi-role analysis and structured triage.

What's the best way to triage code review findings and deduplicate defects?

The best way to triage code review findings is through structured deduplication and classification that prioritizes results into patch, defer, dismiss, or decision-needed outcomes.

How does code review integrate with sprint tracking and story-file updates?

Code review integrates with sprint tracking by updating story files and syncing sprint status, including deferred work tracking with checkpointed human confirmation during the workflow.

Why do I need human confirmation checkpoints during automated code review?

Human confirmation checkpoints are needed during automated code review to validate structured triage outcomes and ensure deferred work tracking and sprint-status updates are accurate before proceeding.