dispatch-middle

Dispatch a bead's middle to isolated subagents with content-bridge artifact relay.

Updated May 3, 2026
One-click install
npx skills add https://github.com/fkberthold/loom --skill dispatch-middle
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dispatch-middle
Source: https://github.com/fkberthold/loom/tree/main/skills/dispatch-middle
Command: npx skills add https://github.com/fkberthold/loom --skill dispatch-middle

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Centralized beads often fight the RED→GREEN middle by bloating context and forcing inline changes. Dispatch-middle splits the middle into independent subagents, each running in its own isolated worktree, so the central orchestrator can dispatch once and never write intermediary state.

Core Features & Use Cases

  • Independent subagents: each middle participant runs in its own worktree with isolated context.
  • Content-bridge relays artifacts: red test content flows between test-author and implementer without shared disks.
  • Lean central: central delegates the middle execution entirely to subagents and only collects results.
  • Use case: non-trivial bead middles where verification and cross-agent collaboration are required.

Quick Start

Invoke /dispatch-middle <bead> to route the RED→GREEN middle through independent subagents.

Frequently Asked Questions about dispatch-middle

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

FAQPage Schema
How do I orchestrate subagents for non-trivial bead RED to GREEN middle execution?

Subagent orchestration for non-trivial bead middles uses dispatch-middle to split execution into isolated worktrees, keeping the central orchestrator lean by delegating the RED→GREEN pipeline entirely to independent subagents.

What is an isolated worktree pipeline for subagent orchestration?

An isolated worktree pipeline gives each subagent its own independent context and workspace, allowing test-author and implementer subagents to run separately while a content-bridge relays red test artifacts between them without shared disks.

How do I run dispatch-middle to route a bead middle through subagents?

Invoke /dispatch-middle <bead> to automatically route the bead's RED→GREEN middle through independent subagents, enforcing isolation, minimal central state, and automatic dispatch sequencing with a default background-run model policy.

How does content-bridge relay artifacts between isolated subagents?

Content-bridge relays red test artifacts between isolated subagents across independent worktrees without shared disks, enabling cross-agent collaboration while the central orchestrator only collects final results.

When should I use subagent orchestration for bead middle execution?

Use subagent orchestration when a bead's RED→GREEN middle is non-trivial and requires verification across isolated worktrees, specifically when centralized execution would bloat context and force unwanted inline intermediary state changes.

Does dispatch-middle enforce a specific run model policy for subagent dispatch?

Dispatch-middle enforces automatic dispatch sequencing with a default background-run model policy, requiring no dependencies while maintaining minimal central state and strict isolation across all subagent worktrees.