orchestrator

Coordinate multi-agent beads across tracks and report outcomes to the epic thread.

225|71|Updated Jan 9, 2026
One-click install
npx skills add https://github.com/hoangnb24/skills --skill orchestrator-hoangnb24
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: orchestrator
Source: https://github.com/hoangnb24/skills/tree/main/.agents/skills/orchestrator
Command: npx skills add https://github.com/hoangnb24/skills --skill orchestrator-hoangnb24

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Orchestrates multiple autonomous agents to execute beads across tracks, ensuring progress is tracked and blockers are surfaced.

Core Features & Use Cases

  • Spawn and manage parallel worker subagents per track based on a provided execution plan
  • Monitor progress and cross-track blockers via Agent Mail, aggregating outcomes for the epic
  • Coordinate escalation and completion reporting back to the epic thread

Example use case: An epic with three tracks can be split into BlueLake, GreenCastle, and RedStone to run beads concurrently and converge on a final deliverable.

Quick Start

Load the execution plan from history/<feature>/execution-plan.md and initialize the orchestrator to spawn track workers.

Frequently Asked Questions about orchestrator

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

FAQPage Schema
How do I coordinate multiple agents to execute tasks in parallel?

To coordinate parallel agents, you can use an orchestrator to spawn worker subagents per track based on an execution plan, monitor their progress via Agent Mail, and aggregate outcomes to complete complex epics.

What is an execution plan and how does it help manage multi-agent epics?

An execution plan defines the tracks and beads for a complex epic. By reading it from history, the orchestrator initializes Agent Mail and spawns workers to run beads concurrently and converge on a final deliverable.

How do I track progress and blockers across parallel agent tracks?

You track cross-track progress and blockers by initializing Agent Mail, which allows the orchestrator to monitor parallel bead execution, surface blockers, and report completion outcomes back to the epic thread.

How do I start an orchestrator for a new epic?

To start a new epic, load the execution plan from history/<feature>/execution-plan.md, initialize the orchestrator to set up Agent Mail, and spawn track workers to begin parallel bead execution.

When should I use multi-agent orchestration for my software engineering tasks?

You should use multi-agent orchestration when starting a new epic that can be split into independent tracks, allowing you to spawn autonomous workers to execute beads concurrently and surface cross-track blockers efficiently.