task-orchestrator

Orchestrate parallel software engineering tasks across isolated git worktrees.

1|Updated Jul 22, 2026

Search Tags:#automation #git #workflow #agent #orchestration #parallel

Ready to run

npx skills add https://github.com/wyattjoh/skills --skill task-orchestrator-wyattjoh

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: task-orchestrator
Source: https://github.com/wyattjoh/skills/tree/main/skills/task-orchestrator
Command: npx skills add https://github.com/wyattjoh/skills --skill task-orchestrator-wyattjoh

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires commander, and includes scripts (resource) components.

What problem does it solve?

This Skill solves the challenge of managing large, multi-step engineering projects by coordinating multiple isolated agents to work in parallel, ensuring reliable integration and consistent progress tracking.

Core Features & Use Cases

Parallel Task Execution: Fan out work units to independent agents, each operating in their own git worktree.
Automated Review Loop: Enforces a rigorous review-fix cycle for every task, ensuring high-quality code before integration.
State Persistence: Maintains a durable state journal that allows the coordinator to resume complex batches after interruptions or session restarts.

Quick Start

Use the task orchestrator to coordinate the work units defined in the plans directory by running the parallelize this work command.

Frequently Asked Questions about task-orchestrator

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

FAQPage Schema

How do I parallelize multi-unit refactoring projects across independent git worktrees?▼

You can parallelize multi-unit refactoring projects by fanning out tasks to isolated agents, each operating in its own git worktree. The coordinator manages their lifecycle through a deterministic dependency graph to ensure reliable integration into a target branch.

How does automated review work for parallel agent workflows?▼

Automated review for parallel agent workflows enforces a rigorous review-fix cycle for every task before integration. This ensures high-quality code by automatically routing agent outputs through validation loops within the deterministic dependency graph.

Can I resume complex engineering workflows after an interruption or session restart?▼

Yes, you can resume complex engineering workflows after interruptions because the coordinator maintains a durable state journal. This state persistence tracks atomic progress, allowing parallelized code changes to restart exactly where they left off.

What is the best way to coordinate multiple agents for feature implementations?▼

The best way to coordinate multiple agents for feature implementations is to use a deterministic dependency graph that manages isolated task execution. This approach applies to parallelizing code changes by serializing the integration of completed units into a target branch.

Do I need commander to manage parallel task execution and orchestration?▼

Yes, you need commander as a dependency to manage parallel task execution and orchestration. It provides the necessary framework for the coordinator to fan out work units to independent agents and manage their lifecycle.