parallel-worktree-orchestrator

Coordinate multiple Claude Code agents across Git worktrees with phase-driven orchestration.

1|Updated Jan 24, 2026
One-click install
npx skills add https://github.com/khalidelborai/claude_skills --skill parallel-worktree-orchestrator
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: parallel-worktree-orchestrator
Source: https://github.com/khalidelborai/claude_skills/tree/main/skills/parallel-worktree-orchestrator
Command: npx skills add https://github.com/khalidelborai/claude_skills --skill parallel-worktree-orchestrator

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires wt, tmux, jq, git, python3, claude, and includes scripts (resource) components.

What problem does it solve?

Orchestrates multiple Claude Code agents across git worktrees to accelerate large-scale implementations by coordinating parallel tasks and consolidating results.

Core Features & Use Cases

  • Distributed agent orchestration across multiple worktrees to speed up long-running code tasks.
  • Phase-driven lifecycle including setup, analysis, launch, monitor, merge, and resume.
  • Centralized session state, dependency management, and deterministic merges via a shared manifest and a suite of helper scripts.

Quick Start

  1. Ensure prerequisites are installed: wt (worktrunk CLI), tmux, git, and the Claude CLI.
  2. For each workstream, create a git worktree, start a Claude Code agent in a tmux session, and monitor progress with the monitor-scripts.
  3. When all agents complete, merge results in order and clean up resources using the provided resume and cleanup workflows.

Frequently Asked Questions about parallel-worktree-orchestrator

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

FAQPage Schema
How do I run parallel Claude Code agents across git worktrees for large-scale implementation?

It coordinates multiple Claude Code agents across git worktrees by applying phase-driven orchestration, establishing non-overlapping task boundaries, and maintaining persistent session state for coordinated launches and deterministic merges.

What dependencies do I need to coordinate parallel workstreams with Claude Code?

You need wt (worktrunk CLI), tmux, jq, git, python3, and the Claude CLI installed to run the parallel agent orchestration. These tools manage the worktrees, terminal multiplexing, and session state required for distributed execution.

How does tmux work with git worktrees to manage multiple AI agents?

Tmux provides isolated terminal sessions for each git worktree, allowing multiple Claude Code agents to run simultaneously without overlapping. The orchestrator uses these sessions to monitor progress and manage the lifecycle of each parallel workstream.

Can I split a complex software project into parallel workstreams using Claude CLI?

Yes, complex software projects can be split into parallel workstreams by establishing non-overlapping task boundaries. The orchestrator manages a shared manifest and helper scripts to handle dependencies and consolidate the workstreams into ordered merges.

What is the best way to merge results from multiple parallel coding agents?

The best way to merge results from parallel coding agents is through a deterministic merge phase driven by a shared manifest. This ensures consolidated results are merged in the correct order after all agents complete their assigned tasks.

How do I resume a Claude Code orchestration session if it gets interrupted?

You can resume interrupted orchestration sessions using the provided resume workflows, which rely on centralized session state and persistent manifests. This allows you to recover progress and continue monitoring or merging the active worktrees.