bmad-review-adversarial-general

Reviews diffs, specs, and documents with adversarial skepticism to produce a findings report.

2|Updated Apr 19, 2026
One-click install
npx skills add https://github.com/Wizarck/nexandro --skill bmad-review-adversarial-general-wizarck
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bmad-review-adversarial-general
Source: https://github.com/Wizarck/nexandro/tree/main/skills/bmad-review-adversarial-general
Command: npx skills add https://github.com/Wizarck/nexandro --skill bmad-review-adversarial-general-wizarck

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Superficial reviews miss hidden defects, gaps, and unstated assumptions in code changes, specifications, and documents. This Skill applies a deliberately cynical, skeptical review process that assumes problems exist and actively hunts for what is missing, not just what is visibly wrong. ## Core Features & Use Cases - Adversarial Analysis: Reviews any artifact (diff, branch, spec, story, or document) with extreme skepticism, guaranteeing at least ten concrete findings. - Optional Focus Areas: Accepts an also_consider input to steer the review toward specific risk areas alongside the general analysis. - Structured Findings Report: Outputs findings as a clean Markdown list with precise, professional descriptions. - Use Case: Before merging a pull request, run this Skill on the diff to surface missing error handling, undocumented assumptions, and edge cases the author overlooked. ## Quick Start Ask the AI to perform an adversarial review of the current diff or a pasted document and list all findings.

Frequently Asked Questions about bmad-review-adversarial-general

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

FAQPage Schema
How do I get a critical review of a code diff before merging?▼

Provide the diff, branch, or uncommitted changes as the content input and request an adversarial review. The Skill analyzes the change with extreme skepticism and returns at least ten findings as a Markdown list.

What types of content can an adversarial review analyze?▼

The review accepts any artifact: code diffs, branches, uncommitted changes, specifications, user stories, or general documents. It automatically identifies the content type before beginning the analysis.

Can I focus the review on specific risk areas?▼

Yes, use the optional also_consider input to specify areas to keep in mind during the review. These are evaluated alongside the standard adversarial analysis rather than replacing it.

What happens if the review finds zero issues?▼

Zero findings is treated as suspicious and triggers a halt condition. The reviewer re-analyzes the content or asks for guidance, since the process assumes problems exist in any submitted work.

When should I not use an adversarial review?▼

Avoid it for early brainstorming drafts where exhaustive fault-finding is premature, or when you need a quick approval check rather than deep scrutiny. It is designed for artifacts ready for rigorous critique.