po-council

# Atlas PO review — two critics, one accountable owner The active pilot does not summon a standing committee. `pnpm po:route` derives the primary risk from change facts and all four boundary assessments, then selects Evidence plus exactly one specialist. The human owner decides. Both seats apply the **Human value** section of `docs/PRODUCT-OWNER-OPERATING-SYSTEM.md`: test the person's before/after ability, not comfort or feature volume. Use the existing outcome and recovery-proof fields for understanding, evidence-calibrated confidence, intervention, and later reuse. Do not introduce another score, reviewer, or approval stage. Give each seat the current **Atlas product thesis** in `docs/PRODUCT-DIRECTION.md` as primary context. The seats are Atlas specialists, not independent generic product consultants: they share the source-meaning, task-brief, human-review, and next-task-reuse model, while independently testing its assumptions from their assigned risk. Preserve evidence against that model. ## Reviewer map These are available roles, not a standing five-person council. Ordinary reversible work stays solo; a routed review uses Evidence and one specialist. `chief` coordinates when needed and is not another review vote. | Derived risk | Required reviewers | Independent question | |---|---|---| | `meaning` | `po-evidence` + `po-steward` | Is the lost human ability observed, and do meaning, evidence truth, local-first behavior, agent authority, and correction stay trustwort...