parallel-orchestratorlisted

<!-- ARCHIVED: This skill has been superseded by orchestrator-directives-skill --> <!-- SDK references removed — use Go CLI commands instead --> # Parallel Orchestrator Skill (ARCHIVED) > **This skill is archived.** Use `/wipnote:orchestrator-directives-skill` for current orchestration patterns. ## Core Principle: 6-Phase Parallel Workflow ``` 1. ANALYZE → wipnote analytics summary + wipnote find features --status todo 2. PREPARE → Cache shared context, isolate tasks 3. DISPATCH → Spawn agents in ONE message (parallel) 4. MONITOR → Track health metrics per agent 5. AGGREGATE → Collect results, detect conflicts 6. VALIDATE → uv run pytest && uv run ruff check ``` --- ## Phase 1: Pre-Flight Analysis ```bash # Check what can be parallelized wipnote analytics summary wipnote find features --status todo wipnote analytics summary ``` ### Decision Criteria | Condition | Action | |-----------|--------| | 2+ independent todo features | Can parallelize | | Shared file edits | Partition or sequence | | All features blocked | Resolve bottlenecks first | --- ## Phase 2: Context Preparation Identify files all agents need and share that context in each agent's prompt. Partition file ownership to avoid conflicts. --- ## Phase 3: Dispatch with Task Tool **CRITICAL: Send ALL Task calls in a SINGLE message for true parallelism!** ```python # CORRECT: All in one message (parallel) Task(subagent_type="wipnote:gemini-operator", description="Research API", prompt="...") Task