swarm-dispatch

Coordinate wave-based swarm task execution with filesystem isolation and merge auditing.

5|Updated Apr 29, 2026
One-click install
npx skills add https://github.com/friedbotstudio/baseline --skill swarm-dispatch
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: swarm-dispatch
Source: https://github.com/friedbotstudio/baseline/tree/main/.claude/skills/swarm-dispatch
Command: npx skills add https://github.com/friedbotstudio/baseline --skill swarm-dispatch

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill orchestrates approved swarm plans so multiple implementation tasks can run in parallel without losing control over scope, order, or safety.

Core Features & Use Cases

  • Wave-by-wave orchestration: Executes tasks in planned waves, updating task status and stopping the run when a failure appears.
  • Filesystem isolation: Uses git worktrees in isolated mode so each worker can modify only its declared write set.
  • Merge auditing: Verifies every changed file against the task contract before applying the patch back to the main branch.
  • Failure recovery: Preserves failed worktrees for inspection and supports clean cleanup when a task makes no changes.
  • Use Case: A lead agent has a multi-task feature plan and wants parallel workers to implement each task safely while maintaining strict write-set discipline.

Quick Start

Use the swarm-dispatch skill to run the approved swarm slug from .claude/state/swarm and guide each wave through isolated worker execution and merge audit.

Frequently Asked Questions about swarm-dispatch

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

FAQPage Schema
How do I run parallel git worktree tasks safely without merge conflicts?

Parallel git worktree tasks run safely through wave-based orchestration that enforces filesystem isolation, per-worker write-set discipline, and baseline diff auditing before applying patches back to the main branch.

What is swarm task orchestration and when do I need it for multi-task implementation?

Swarm task orchestration coordinates approved multi-task implementation plans by executing tasks in planned waves with parallel workers, updating JSON status reporting, and stopping the run when a failure appears.

How to execute wave-based parallel workers and audit merges back to the main branch?

Execute wave-based parallel workers by applying isolated git worktrees to each task, then verify every changed file against the task contract through baseline diff auditing before merging the patch back to the main branch.

Does git worktree filesystem isolation prevent workers from modifying undeclared files?

Git worktree filesystem isolation restricts each worker to modifying only its declared write set, enforcing strict write-set discipline and preventing unauthorized file changes during parallel task execution.

What happens to failed worktrees when parallel task execution stops a wave?

When a failure appears, failure-aware wave stopping halts the run and preserves failed worktrees for inspection, while supporting clean cleanup when a task makes no changes to the repository.

Can I use swarm task orchestration for multi-task feature plans without losing control over scope?

Swarm task orchestration executes approved multi-task feature plans with parallel workers while maintaining strict write-set discipline, controlled handoff back to the main context, and enforced task contract verification.