Triage

Every feature across every system, one row each. Filter, sort, and use this to work the backlog — then record the team's actual call in status.json's team_status field per feature (the stamp shown here is Claude's suggestion, not a decision).