teamwork

Orchestrate multi-agent tasks with parent-led coordination and MCP worktree management.

Updated Mar 31, 2026
One-click install
npx skills add https://github.com/moorej2400/j-skills-public --skill teamwork
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: teamwork
Source: https://github.com/moorej2400/j-skills-public/tree/main/teamwork
Command: npx skills add https://github.com/moorej2400/j-skills-public --skill teamwork

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Coordinated, parent-led orchestration of tasks across multiple specialized agents working in isolated worktrees, reducing drift and speeding delivery.

Core Features & Use Cases

  • Phase-based decomposition: break a task into phases with dedicated workers and shared integration.
  • Lightweight MCP coordination: enables session, work-item, and status management between parent and workers.
  • Worktree registration and integration: tracks worker worktrees and merges results into main with traceable logs.

Quick Start

Review the teamwork task plan, configure defaults in teamwork/.env, and start the parent-led MCP workflow to coordinate specialists.

Frequently Asked Questions about teamwork

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

FAQPage Schema
How do I coordinate multi-agent tasks across isolated worktrees without drift?

Multi-agent coordination across isolated worktrees is managed through a parent-led orchestration model using a dedicated teamwork-mcp server for worker registration, phase checkpoints, and result integration.

Can I orchestrate multi-agent tasks in solo or pair modes instead of using all six workers?

Yes, multi-agent orchestration supports solo or pair modes in addition to coordinating up to six workers, allowing flexible phase-based decomposition based on task complexity.

What is phase-based decomposition for multi-agent orchestration?

Phase-based decomposition breaks a task into sequential phases with dedicated workers and shared integration, managed through lightweight MCP coordination to track work items and status between parent and workers.

How do I merge multi-agent worker results into main with traceable logs?

Merging worker results into main is handled by worktree registration and integration tracking, which logs each merge operation through the teamwork-mcp server to ensure traceable delivery.

Do I need a dedicated MCP server to manage multi-agent worktree registration?

Yes, a dedicated teamwork-mcp server is required to manage session, work-item, and status coordination between the parent and workers, enabling lightweight worktree registration and phase checkpoints.

What are the limitations of parent-led multi-agent orchestration?

Parent-led orchestration is limited to coordinating up to six workers and requires phase-based decomposition; tasks outside this worker capacity or lacking clear phase boundaries may not suit this model.