What problem does it solve? Plans and design proposals often contain hidden assumptions, unexamined failure modes, and blind spots that only surface after resources are committed. This Skill stress-tests any plan before execution, surfacing blockers, risks, and gaps while changes are still cheap. ## Core Features & Use Cases - Structured Multi-Lens Review: Applies seven lenses — first principles, gaps and blind spots, edge cases and scale, alternatives, implementation risk, press release test, and counter-plan — plus pre-mortem and Murphyjitsu calibration. - Multi-Reviewer Dispatch Mode: Spawns parallel critic subagents (Architect, Skeptic, Operator, User, Counter-Planner) with anti-groupthink guards like the 10th Man Rule for major initiatives. - Classified Findings Output: Produces a structured review with Blockers, Risks, Gaps, Questions, an assumption map, and a summary table, delivered in chat or as GitHub Issues. - Use Case: Before committing a team to a six-week migration plan, run an adversarial review to surface the unstated dependency on a legacy service and the missing rollback strategy, then revise the plan before implementation starts. ## Quick Start Ask the AI to run an adversarial review on your plan document, for example by pointing it at a plan file path and requesting a stress-test of its assumptions and failure modes.
