task-handoff
Task Handoff
State file: ~/.openclaw/skill-state/task-handoff/state.yaml
Pre-Handoff Checklist
- Current state is stable (no half-written files, no broken tests)
- All changed files are saved
- Tests have been run
- Handoff document is written
- State file is updated
- Memory is updated
Writing a Handoff
Save to tasks/handoff-[task-name]-[timestamp].md:
# Handoff: [task name]
Written: YYYY-MM-DD HH:MM
Reason: [why stopping]
## Current State
[1-3 sentences: exactly where things are]
## What's Done
- [step] done
## What's Next
- [next step] - specific instructions
## Important Context
- [anything non-obvious from the code]
- [decisions made and why]
## Files Modified
- path/to/file - [what changed]
## Tests
- Last run: YYYY-MM-DD HH:MM
- Status: PASSING / FAILING
## Blockers
- [anything blocking]
Then update state: active_handoff_path (full path to the file), task_name, reason, status: written, written_at, files_modified.
Picking Up a Handoff
- Read state file to get
active_handoff_path - Read the handoff document completely before touching any code
- Run the tests to confirm current state
- Start from 'What's Next' — do not redo completed steps
- When complete: delete handoff file, update state
status: picked_up, clearactive_handoff_path
More from archieindian/openclaw-superpowers
context-window-management
Prevents context overflow on long-running OpenClaw sessions. Use when approaching context limits.
26heartbeat-governor
Enforces per-skill execution budgets for scheduled cron skills — pauses runaway skills that exceed their token or wall-clock budget before they drain your monthly API allowance.
25using-superpowers
Bootstrap skill — teaches the agent how to find and invoke skills. Use when starting any new task or session.
20long-running-task-management
Breaks multi-hour tasks into checkpointed stages with resume capability. Use when a task is expected to take more than 30 minutes or multiple sessions.
19fact-check-before-trust
Triggers a secondary verification pass for any agent output containing factual claims, numbers, dates, or named entities before the output is acted on
19agent-self-recovery
Detects when the agent is stuck in a loop and escapes systematically. Use when you notice repeated failures or loss of direction.
18