avocat-du-diable

Identifies and challenges plan, architecture, and code decisions with steel-man critique and structured questioning.

1|Updated Apr 5, 2026
One-click install
npx skills add https://github.com/Alexmacapple/alex-claude-skill --skill avocat-du-diable
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: avocat-du-diable
Source: https://github.com/Alexmacapple/alex-claude-skill/tree/main/avocat-du-diable
Command: npx skills add https://github.com/Alexmacapple/alex-claude-skill --skill avocat-du-diable

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Structured critical review of AI work (plans, code, architecture) using steel-man reasoning, questioning frameworks, and a concrete verdict to prevent premature consensus.

Core Features & Use Cases

  • Autonomous and coupled review modes to review outputs or prior artifacts.
  • Four-step review process: identify target, steel-man, apply questioning cadres, deliver a clear verdict.
  • References loading and guidance on handling plans, architecture, and code artifacts.

Quick Start

Dites-moi ce que vous souhaitez que je révise et fournissez le fichier ou le plan à analyser.

Frequently Asked Questions about avocat-du-diable

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

FAQPage Schema
What is a steel-man critique and how does it work for code review?

A steel-man critique involves identifying the target decision plan, constructing the strongest possible opposing argument, applying structured questioning frameworks for risk-analysis, and delivering a concrete verdict to prevent premature consensus.

Does code review risk-analysis support autonomous and coupled review modes?

Yes, you can apply structured questioning to review both code and architecture artifacts. The process supports autonomous review of generated outputs and coupled review of prior artifacts to flag blocking concerns.

What's the best way to get an actionable verdict from a plan critique?

To start a critical review of decision-making plans, provide the plan or file you want analyzed and request a review. The system will identify the target, apply questioning, and deliver a clear verdict with flagged concerns.

When should I not use a steel-man approach for decision-making reviews?

The primary limitation of using steel-man reasoning for risk-analysis is that it focuses strictly on structured critical review of plans, architectures, and code, requiring a defined target artifact to initiate the four-step critique process.