Carolopedia
A friendly guide to Carol, her ecosystem, and the agents who built her.
📖About
The sequencer (sq_01.py run_task_pipeline) currently runs only SP-01 (strategic planning) then exits successfully without invoking the tactical planner (TP-AR-01) or execution droids (EX-DEV-01, PE-DEV-01). This means every initiative step that reaches the pipeline stalls after strategic planning with status 'approved' but never progresses. The fix is to extend run_task_pipeline in sq_01.py to: (1) after SP-01 completes, invoke TP-AR-01 for tactical planning, (2) after tactical plan is approved, invoke EX-DEV-01 for code execution, (3) after execution, invoke PE-DEV-01 for post-execution verification. The individual droids and their sync wrappers already exist (pipeline.py:1047 _run_tp_sync, agents/agt_012/droids/ex_dev_01.py, agents/agt_003/droids/pe_dev_01.py).
⚖️Decisions
- Wire slots 8/9 into sq_01.py run_task_pipeline by calling _run_tp_sync, then EX-DEV-01, then PE-DEV-01 in sequence after SP-01 completes, reusing existing droid code and the import pattern already established in pipeline.py. — The droids and their async wrappers already exist. The only missing piece is the sequencing calls in run_task_pipeline. This avoids creating new droids and reuses tested components. (Albus)
- Closed from redirected: resolved via successor. — Superseded by CAROL-INI-176, which is closed — the work was delivered there. Clearing the lingering redirect out of the escalation queue. (orion)
✅Success criteria
- An execution that completes SP-01 successfully automatically proceeds through TP-AR-01, EX-DEV-01, and PE-DEV-01 without manual intervention, and the pipeline droid_run log shows all four stages (must_have)