What is a 'Steel-man' argument in this context?

A steel-man argument involves Claude presenting the strongest possible version of your current choice before critiquing it, ensuring the analysis is fair and addresses the best possible version of your logic.

When should I activate the Devil's Advocate skill?

You should use it when choosing a new technology stack, making architectural decisions, or after generating documentation to ensure no critical requirements were missed.

How does the 'Minimum Gaps Rule' work?

The skill is programmed to identify at least two potential gaps or concerns per document. This prevents 'rubber-stamping' and forces the AI to actively look for missing requirements or failure modes.

Does this skill provide actionable advice?

Yes, every analysis concludes with a clear, actionable verdict and recommendation to ensure you aren't left in analysis paralysis.

Devil's Advocate Analysis

Name: Devil's Advocate Analysis
Author: igorder-dev

byigorder-dev

0•

Learning & Documentation

Performs critical stress-tests on technical decisions and documentation to identify hidden risks and architectural gaps.

This skill empowers Claude to act as a constructive contrarian, systematically challenging technical assumptions, architectural choices, and documentation completeness. By employing a three-level analysis—decision critique, scope evaluation, and a 'first-day' usability test—it ensures that projects are not just functional but resilient and well-documented. It enforces a 'Minimum Gaps Rule' to ensure no stone is left unturned, providing actionable recommendations to move projects from raw ideas to production-ready status with confidence.

Key Features

01Minimum Gaps Rule that forces identification of at least two potential concerns per doc

02Automatic scope critique to identify forgotten requirements or failure modes

03First Day on the Job test to ensure documentation clarity for new developers

040 GitHub stars

05Steel-man analysis to evaluate the strongest version of current arguments before critique

06Multi-level risk assessment covering probability, impact, and detectability

Use Cases

01Stress-testing architectural decisions before finalizing Architecture Decision Records (ADRs)

02Evaluating build-vs-buy tradeoffs for frameworks or third-party service integrations

03Reviewing generated documentation for missing edge cases, security concerns, or implicit requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add igorder-dev/apex devil-advocate

For use in Claude.ai and ChatGPT

Download Skill