critical-review

Identify bugs, design flaws, and logical errors in code and text reviews.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/mickaelyoshua/dev --skill critical-review
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: critical-review
Source: https://github.com/mickaelyoshua/dev/tree/main/.claude-skills/critical-review
Command: npx skills add https://github.com/mickaelyoshua/dev --skill critical-review

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Finds bugs, design flaws, and logical errors through adversarial code and text review.

Core Features & Use Cases

  • Systematically identifies critical issues in code and documentation
  • Ranks findings by severity (CRITICAL, MAJOR, MINOR) with concrete references
  • Provides actionable remediation guidance and evidence-backed feedback

Quick Start

Initiate an adversarial review by providing the target code or document and request a structured critique.

Frequently Asked Questions about critical-review

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

FAQPage Schema
How do I perform an adversarial code review to find security vulnerabilities?

Adversarial code review systematically identifies bugs, design flaws, and logical errors by challenging assumptions in code, categorizing findings by severity, and outputting structured remediation guidance.

What is the best way to audit technical documentation for design flaws?

Auditing technical documentation for design flaws involves applying adversarial review to the text, ranking identified issues by severity, and providing evidence-backed feedback with concrete passage references.

Can I use this for quality assurance on both code and text reviews?

Yes, you can use it for quality assurance across software projects; it accepts both code and technical documents to produce categorized findings and actionable remediation guidance.

How do I get structured remediation guidance for identified code bugs?

To get structured remediation guidance for code bugs, initiate an adversarial review by providing the target code; the system outputs categorized findings with concrete line references and actionable fixes.

Does the code review output rank issues by severity levels?

Yes, the code review output ranks issues by severity levels, categorizing them into CRITICAL, MAJOR, and MINOR findings while referencing concrete lines or passages for evidence-backed feedback.

When should I not rely on automated adversarial review for finding logical errors?

You should not rely solely on automated adversarial review when you lack the prerequisite context of the software project's design intentions, as finding complex logical errors requires understanding the intended system behavior.