Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Dispatches multiple subagents to work on independent tasks at the same time instead of sequentially. Each subagent receives a focused, self-contained prompt scoped to one domain. Suited for situations where tasks have no shared state or ordering dependencies.
Sequential task handling multiplies total time by the number of tasks, while parallel dispatch collapses that to roughly the duration of the longest single task.
When to use it
- Fixing failures in three unrelated test files at once
- Running a UI bug fix and an API optimization simultaneously
- Parallelizing research, implementation, and test writing across separate agents
- Debugging multiple independent subsystem failures in one session
- Generating assets while another agent refactors code
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Fix the collision detection bug in physics.test.ts, optimize the render pipeline, and write tests for the new level loader
- 1Identify the three tasks as independent domains with no shared state
- 2Write a focused, self-contained prompt for each task including file paths and expected output
- 3Dispatch three subagents simultaneously via sessions_spawn
- 4Wait for all agents to return their summaries
- 5Review each result for conflicts, run full test suite, and spot-check changes
Three completed workstreams merged: bug fixed, pipeline optimized, tests written, all passing together