What problem does it solve? Implementation plans often contain hidden assumptions, scope creep, missing edge cases, and wasted effort that only surface during execution. This Skill audits a drafted plan from multiple adversarial perspectives before any code is written, catching fatal flaws early. ## Core Features & Use Cases - Multi-Agent Council Review: Dispatches parallel agents including a Skeptic, Pragmatist, Temporal analyst, Backward Inductor, Scope Creep Finder, Gap Finder, Library Awareness agent, Cost Auditor, and Non-Reliance Auditor to attack the plan from different angles. - Structured Synthesis: Resolves conflicts between agents, identifies consensus findings and blind spots, and produces a verdict of Proceed, Proceed with modifications, or Rethink. - Scope Governance: Applies a 48-hour rule and budget-based trade-offs so audit findings do not inflate scope. - Use Case: After drafting a plan.md for a new feature, run the audit to discover that task 4 assumes an API behavior task 2 never establishes, and that two tasks reimplement functionality an existing library already provides. ## Quick Start Ask the agent to audit the plan.md file in the current working directory using the plan audit council before implementation begins.
