What problem does it solve?
This Skill eliminates the dangerous vagueness of design anxiety by breaking it into concrete attack angles, defending against them, and explicitly stating residual risks before a project reaches completion.
Core Features & Use Cases
- Multi-Wave Adversarial Review: Runs challenger attacks across six mandatory angles including reason for existence, real-use verification, bus factor, platform obsolescence, self-referential structure, and gate-locality.
- AI-Specific Threat Modeling: Meta-Aware Adversary mode targets AI-specific risks like hallucination, context collapse, prompt injection, and tool lock-in.
- Gate-Passage Re-Attack: Wave-P3 re-attacks artifacts after upstream gates declare PASS, checking coverage gaps, narrative residue, and false confidence.
- Use Case: Imagine you have a near-complete design document. Use this Skill to run an all-angle review that surfaces critical flaws, defends against them, and produces a residual risk card before the team commits to implementation.
Quick Start
Use the steel-quench skill to run an all-angle verification on the attached near-complete artifact and produce a residual risk card.
