threnody-subtaskslisted

# Threnody subtask monitoring `list_subtasks` tracks **`execute_subtask`** utility delegation only. Host-native swarms and `host_spawn_waves` do **not** appear here — track those by wave completion in the host shell. Token-cheap review and swarm changes happen on the host-native wave path, not in this utility subprocess monitor. Use this skill only after explicit `execute_subtask` delegation to utilities. ## When to use - User enabled `providers.delegation_utilities_enabled: true` - Work was delegated via `execute_subtask` to OpenCode, Aider, or a local endpoint - Need to pause/resume a running utility subprocess ## Process 1. Call **`list_subtasks`** (MCP: Threnody). 2. Format **Running Active** (`active_count` total): For each group in `active_groups`: Parallel wave (`parallel: true`): ``` WAVE: wave-1 (3 running in parallel) - abc123 [low] gpt-5-mini 12.3s -> config.py - def456 [low] gpt-5-mini 12.1s -> models.py ``` Single task: ``` abc123 [med] claude-sonnet-4.6 4.2s -> main.py ``` Each row: `task_id` (first 6 chars), `[tier]`, model or `resolving`, elapsed seconds, `target_file` or prompt excerpt (60 chars). If `active_count == 0`: *No subtasks currently running.* 3. Format **Recently Completed** (last 10): ``` abc123 DONE [low] gpt-5-mini / Aider 28.1s -> config.py def456 FAILED [low] unknown 31.0s -> models.py ``` Group consecutive entries with the same `wave_id` under a WAVE label. 4. Footer: `X active / Y