bmad-code-review

Review code changes with parallel adversarial layers and structured triage.

Updated Mar 16, 2026

Search Tags:#testing #developer tools #quality assurance #code review #triage #adversarial testing

Ready to run

npx skills add https://github.com/deltatree/wissensmanagement --skill bmad-code-review-deltatree

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bmad-code-review
Source: https://github.com/deltatree/wissensmanagement/tree/main/_bmad/bmm/workflows/4-implementation/bmad-code-review
Command: npx skills add https://github.com/deltatree/wissensmanagement --skill bmad-code-review-deltatree

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

This Skill automates the process of reviewing code changes by employing multiple adversarial review layers to identify potential issues, ensuring higher code quality and adherence to specifications.

Core Features & Use Cases

Parallel Review Layers: Utilizes distinct review strategies (Blind Hunter, Edge Case Hunter, Acceptance Auditor) to cover various aspects of code quality.
Structured Triage: Categorizes identified issues into actionable buckets (intent gap, bad spec, patch, defer) for clear prioritization.
Use Case: When a developer submits a pull request, this Skill can be invoked to perform an automated, multi-faceted review, flagging potential bugs, spec deviations, or areas needing clarification before human review.

Quick Start

Run code review for the current changes.

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 improve quality assurance?▼

Adversarial code review improves quality assurance by deploying parallel review strategies, such as a general hunter and an edge case hunter, to actively identify potential bugs and specification deviations before human review.

How do I triage code review findings into actionable categories?▼

You triage code review findings by categorizing identified issues into structured buckets like intent gaps, bad specifications, patches, or deferrals, ensuring clear prioritization for subsequent development actions.

How do I automate pull request reviews for edge cases and spec deviations?▼

Automate pull request reviews by invoking a multi-faceted review process that runs parallel layers, including an acceptance auditor and an edge case hunter, to flag bugs and spec deviations automatically.

Do I need a specific architecture to run automated code reviews?▼

Yes, executing automated code reviews requires a step-file architecture to support disciplined execution and state tracking throughout the multiple adversarial review layers.

What is the best way to identify intent gaps in code changes?▼

The best way to identify intent gaps in code changes is using an adversarial review process that applies an acceptance auditor to evaluate whether the code meets the original specifications and intent.