devil

Generates severity-tagged adversarial critiques of claims or decisions to identify risks and failure modes.

10|1|Updated Jun 29, 2026
One-click install
npx skills add https://github.com/mishahanin/heading-os --skill devil-mishahanin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: devil
Source: https://github.com/mishahanin/heading-os/tree/main/.claude/skills/devil
Command: npx skills add https://github.com/mishahanin/heading-os --skill devil-mishahanin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill eliminates sycophancy and groupthink by forcing a rigorous, adversarial critique of your decisions, claims, or recommendations.

Core Features & Use Cases

  • Adversarial Critique: Generates N severity-tagged (BLOCKER to LOW) points attacking a target from distinct angles like risk, cost, and scope.
  • Honesty Floor: Prevents the fabrication of weak points by stopping early if the AI cannot find substantive, distinct criticisms.
  • Use Case: Before finalizing a strategic pivot or a high-stakes email, use this to identify hidden failure modes or overlooked stakeholder risks.

Quick Start

Invoke the devil skill to provide five critical points against your latest decision by typing /devil 5.

Frequently Asked Questions about devil

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

FAQPage Schema
How do I pressure-test a strategic decision for hidden risks?

Pressure-test a strategic decision by generating structured adversarial critiques that identify potential failure modes across multiple analytical dimensions, providing objective severity-tagged feedback against your specific claims.

Can I use adversarial critique to break groupthink during risk assessment?

Adversarial critique breaks groupthink by forcing rigorous attacks on your claims from distinct angles like cost and scope, maintaining an honesty floor that prevents fabricated weak points and sycophantic output.

How many critical points can I generate when evaluating a high-stakes claim?

Evaluating a high-stakes claim allows you to generate N severity-tagged critical points, such as invoking five distinct critiques, to target your decision from various analytical angles.

Does the critique mechanism require external tools to identify failure modes?

The critique mechanism requires no external tool execution to identify failure modes, relying entirely on internal conversational context logic to evaluate claims and maintain an honesty floor.

What are the limitations of using automated critique for decision-making?

A limitation of automated critique is that it stops early if it cannot find substantive, distinct criticisms, preventing the fabrication of weak points but potentially missing edge cases outside conversational context.