orchestrator-control-skill

Coordinate multi-agent PR lifecycle management across isolated worktrees.

Updated Dec 24, 2025
One-click install
npx skills add https://github.com/hanibalsk/claude-threads --skill orchestrator-control-skill
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrator-control-skill
Source: https://github.com/hanibalsk/claude-threads/tree/main/skills/orchestrator-control
Command: npx skills add https://github.com/hanibalsk/claude-threads --skill orchestrator-control-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a centralized orchestrator control layer to manage PR lifecycles, PR shepherd agents, and multi-agent coordination across worktrees, reducing manual orchestration and scale challenges.

Core Features & Use Cases

  • Master control thread: Spawn and manage the orchestrator daemon lifecycle.
  • PR lifecycle management: Monitor PRs with automatic merge conflict resolution and review handling.
  • Multi-agent spawning: Create per-PR shepherds and sub-agents for conflicts and comments.
  • Isolated worktrees: Every agent runs in its own worktree to avoid contention.

Quick Start

Start the orchestrator control thread and begin watching PRs with auto-merge.

Frequently Asked Questions about orchestrator-control-skill

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

FAQPage Schema
How do I coordinate multiple PR agents across isolated worktrees?

Multi-agent PR coordination uses a master orchestrator thread that spawns per-PR shepherd agents in isolated worktrees, preventing contention and enabling autonomous handling of conflicts, reviews, and merges across parallel workflows.

Can I automate merge conflict resolution in pull requests?

Yes, the orchestrator spawns dedicated conflict resolver sub-agents that autonomously detect and handle merge conflicts, integrated with configurable auto-merge criteria to streamline PR lifecycle management without manual intervention.

What's the best way to manage PR lifecycles with multi-agent automation?

Use a centralized orchestrator control layer to monitor PR health, route review comments to handler sub-agents, spawn shepherd agents per PR, and coordinate inter-agent messaging for escalations—reducing manual orchestration at scale.

How does the orchestrator daemon manage PR shepherd agents?

The master control thread spawns and monitors per-PR shepherds that run in isolated git worktrees, handling lifecycle events, comment workflows, and conflict resolution while maintaining independent execution environments.

Do I need separate agents for different PR workflow stages?

The architecture uses specialized sub-agents: shepherds manage individual PR lifecycles, conflict resolvers handle merge disputes, and comment handlers process review feedback—all coordinated by the central orchestrator for comprehensive workflow coverage.

What prerequisites are needed before running the orchestrator?

Set up git worktrees for isolated agent execution, configure auto-merge criteria and review-comment workflows, and prepare a master control thread environment with inter-agent messaging capability for spawning and managing daemon lifecycle.