Pipeline Status Dashboard

Data Collection (parallel)

1. •Pipeline state: .state/pipeline.json → phase, stage, currentStory, lastAgent, lastRunId, updatedAt
2. •Stories: Glob prd/US-*.json → extract id, title, .stage (top-level), priority, riskTier, sizing, status.deploy.attempts
3. •Areas: .plan/areas.md → parse table for name, status, priority, criticality, open_questions
4. •Planning docs: Check existence of 8 required docs in .plan/: areas.md, work-breakdown.md, traceability.md, runbook.md, decisions.md, assumptions.md, dependencies.md, risk-register.md
5. •Lock: Check .state/.pipeline.lock directory exists

Edge cases: No pipeline.json → "Not initialized, run bootstrap-plan.sh". No stories → skip implementation. No areas.md → skip planning.

Output

=== Evoloop Pipeline Status ===
Updated: <updatedAt>  |  Lock: <Active/None>

--- Pipeline ---
Phase: <phase>  |  Stage: <stage>
Current Story: <currentStory>  |  Last Agent: <lastAgent>  |  Last Run: <lastRunId>

--- Planning Progress ---
Documents: <N>/8 complete  [x] areas.md  [ ] assumptions.md ...
Areas: <N approved/locked>/<total>
  | Area | Status | Priority | Criticality | Open Questions |

--- Implementation Progress ---
Stories: <complete>/<total>
  | ID | Title | Stage | Risk | Size | Deploy Attempts |
By stage: build: N | review: N | deploy: N | complete: N | blocked: N

--- Blockers ---
<Blocked stories with deploy.notes or "No blockers">

--- Next Action ---

Next Action Logic

• •Not initialized → ./scripts/bootstrap-plan.sh
• •Docs incomplete → ./orchestrator.sh plan start --tool claude
• •Areas unapproved → ./orchestrator.sh plan area --area <name> --tool claude
• •No stories → ./orchestrator.sh plan pm --tool claude
• •Incomplete stories → ./orchestrator.sh run --tool claude
• •Blocked → /reset-story <ID> or /inspect <ID>
• •All complete → "Pipeline finished."
• •Lock active → "Pipeline running. Check /logs"