cw-dispatch

Dispatch independent tasks to parallel agent workers for concurrent execution.

14|1|Updated Jan 25, 2026
One-click install
npx skills add https://github.com/sighup/claude-workflow --skill cw-dispatch
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cw-dispatch
Source: https://github.com/sighup/claude-workflow/tree/main/skills/cw-dispatch
Command: npx skills add https://github.com/sighup/claude-workflow --skill cw-dispatch

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill coordinates parallel task execution by identifying independent tasks and spawning dedicated workers after cw-plan to speed up delivery.

Core Features & Use Cases

  • Parallel task dispatch: Group ready tasks by dependency and assign them to separate workers for concurrent execution.
  • Ownership and governance: Ensure tasks are claimed by a single worker and progress is tracked through the workflow.
  • Use Case: When a feature plan yields multiple unblocked tasks, cw-dispatch runs them in parallel to shorten cycle time and improve throughput.

Quick Start

Initiate cw-dispatch after cw-plan to assign ready tasks to worker agents for parallel execution.

Frequently Asked Questions about cw-dispatch

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I speed up task execution by running independent tasks in parallel?

To speed up task execution, you can dispatch independent tasks to parallel agent workers, which accelerates delivery by running unblocked tasks concurrently without shared mutations.

How does parallel task dispatch handle task ownership and dependencies?

Parallel task dispatch relies on TaskList and TaskUpdate to ensure tasks are claimed by a single worker, enforcing dependency ordering and tracking ownership to ensure correct, auditable progress.

Do I need a prior workflow plan to dispatch tasks to parallel workers?

Yes, you need to initiate dispatch after a feature plan yields multiple unblocked tasks. The workflow groups ready tasks by dependency and assigns them to separate workers for concurrent execution.

When should I avoid using parallel workers for task management?

You should avoid using parallel workers for task management when your tasks require shared mutations, as this dispatch mechanism is designed specifically for independent tasks without shared state dependencies.

Can I use task-management workflows to track worker progress and report results?

Yes, the task-management workflow relies on TaskList, TaskUpdate, and worker invoke cw-execute to track ownership, enforce dependency ordering, and report progress to ensure auditable results.