orchestrate

Coordinate Claude Code agents in tmux windows with worktree recycling.

Updated Jan 16, 2026
One-click install
npx skills add https://github.com/ghpworkuse-web/train --skill orchestrate-ghpworkuse-web
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrate
Source: https://github.com/ghpworkuse-web/train/tree/main/AutoGPT/.claude/skills/orchestrate
Command: npx skills add https://github.com/ghpworkuse-web/train --skill orchestrate-ghpworkuse-web

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires jq, tmux, git, uuidgen, python3, gh, curl, claude, perl, sed, and includes scripts (resource) components.

What problem does it solve?

Orchestrate provides a self-contained framework to supervise a fleet of Claude Code agents running in tmux windows. It automatically discovers spare worktrees, spawns agents, monitors state, reboots idle agents, approves safe confirmations, and recycles worktrees when tasks are completed, enabling coordinated multi-agent workflows.

Core Features & Use Cases

  • Auto-discovers spare worktrees and assigns tasks to each agent
  • Spawns agents, tracks per-agent state, and maintains a centralized JSON state file
  • Monitors agent health, restarts idle agents, and prompts for safe approvals
  • Verifies task completion via checkpoints and integration/tests, then recycles worktrees
  • Supports status and capacity checks, capacity planning, and parallel task orchestration
  • Integrates with PR workflows (addressing, testing, and final evaluation) in a controlled, serialized manner

Quick Start

Spawn new agents on available spare worktrees and monitor their progress.

Frequently Asked Questions about orchestrate

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

FAQPage Schema
How do I coordinate multiple Claude Code agents to run in parallel?

Coordinating multiple Claude Code agents involves using a supervisor to manage a fleet running in tmux windows. The system automatically discovers spare git worktrees, spawns agents into them, and maintains a centralized JSON state file to track progress.

How does tmux work with Claude Code for multi-agent orchestration?

Tmux provides the windowing environment where each Claude Code agent runs individually. The orchestration framework spawns agents into separate tmux windows, monitors their health, restarts idle agents, and handles safe confirmation approvals automatically.

How do I manage git worktrees for parallel AI coding agents?

The orchestration framework auto-discovers spare git worktrees and assigns individual tasks to each agent. When tasks are completed and verified via integration tests, the system automatically recycles those worktrees for future use.

Do I need specific tools installed to run Claude Code agent orchestration?

Yes, agent orchestration requires several dependencies including tmux, git, jq, python3, gh, curl, perl, sed, and uuidgen. These tools support state management, git operations, API interactions, and terminal multiplexing.

Can I integrate parallel agent workflows with GitHub PR processes?

Yes, orchestration integrates with PR workflows by addressing, testing, and evaluating pull requests in a controlled, serialized manner. The supervisor verifies task completion via checkpoints and tests before recycling worktrees.

What are the limitations of using tmux for Claude Code agent orchestration?

Using tmux for orchestration requires all dependencies to be correctly installed and configured. The system relies on serialized PR workflows and centralized JSON state files, meaning network failures or missing dependencies can disrupt agent monitoring and worktree recycling.