Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Add watchdog auto-continue mode (30m checks), stalled-task auto-resume behavior, and delivery reliability guidance for main/systemEvent continuity updates.
What This Skill Does
Tracks active tasks and queues them automatically when an unrelated request interrupts progress. Uses a shared JSON queue file and a Python helper script to persist task context across sessions and clones. After the active task finishes, the oldest queued item resumes immediately with full context.
Unlike manual reminders, it persists structured context (what was done, exact next step, acceptance criteria) at interruption time, so resumption requires no re-orientation.
When to use it
- Resuming a half-finished code review after handling an urgent bug fix
- Picking up a paused writing task after answering an unrelated question
- Enforcing finish-first order when side requests arrive mid-migration
- Recovering an interrupted multi-step setup without losing the next-step context
- Running a daily queue check to surface and reschedule stale tasks
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: 'Can you quickly draft an email?' while an unfinished database migration task is active
- 1Detects active unfinished task and incoming unrelated request
- 2Immediately runs task_resume_queue.py add with migration title, progress context, and acceptance criteria
- 3Briefly acknowledges the queue action to the user
- 4Handles the email draft request to completion
- 5Runs task_resume_queue.py pop and resumes the migration from the saved checkpoint
Email drafted, then agent announces 'Resuming previously interrupted task: database migration' and continues from the exact saved step