What problem does it solve? Coordinating a rigorous multi-reviewer audit of a product, codebase, architecture, or GTM artifact is hard to do consistently: findings lack evidence, reviewer outputs contaminate each other, and final reports miss required structure. This Skill turns a scaffolded ARC run skeleton into a complete, schema-validated review bundle. ## Core Features & Use Cases - Independent Reviewer Passes: Runs specialist, skeptic/red-team, and adjudicator passes in isolation so findings stay unbiased before synthesis. - Evidence-Gated Findings: Requires every finding to carry source locators, severity, confidence, and a validation method; unsupported claims stay hypotheses. - Schema-Valid Artifact Bundle: Produces evidence_pack.md, findings.yaml, scorecard.json, risk_register.yaml, decision_record.md, and validation_plan.md under runs/<date>-<slug>/, closed with uv run arc validate. - Use Case: After scaffolding a run with the council-run skill, ask the agent to review a new API design; it collects evidence, runs reviewer passes, adjudicates accepted/rejected/disputed/watchlist findings, and delivers a validated decision record. ## Quick Start Ask the agent to run a council review over the scaffolded run in runs/2026-06-22-api-design and produce the validated artifact bundle.
