prompt logo

prompt

Free

Central dispatch agent for multi-agent orchestration

FreeFree tier
Type
Open Source

About prompt

The Multi-Agent Orchestrator (MAO) system prompt is a comprehensive guide for an AI agent acting as a central dispatcher. It decomposes complex tasks into atomic sub-tasks and delegates them to specialized sub-agents such as researcher, coder, reviewer, data_analyst, and writer. The prompt defines a strict role for the orchestrator (router/coordinator, not executor), a task decomposition protocol, delegation rules for parallel or sequential execution, and a chaining pattern to pass outputs between agents. It is designed to be used within a multi-agent architecture to plan, route, track, and synthesize results.

Key Features

Decomposes complex tasks into atomic sub-tasks
Delegates to specialized sub-agents (researcher, coder, reviewer, data_analyst, writer)
Supports parallel and sequential execution
Chaining pattern for passing outputs between agents
Tracks state of sub-tasks (pending, in-progress, completed, failed)
Read-only tools for context gathering
Synthesizes sub-agent outputs into final result

Pros & Cons

Pros
  • Clear separation of concerns: orchestrator does not execute, only coordinates
  • Supports both parallel and sequential task execution
  • Explicit chaining pattern for dependency management
  • State tracking for sub-tasks
  • Flexible agent definitions via trigger keywords
  • Open source and free to use
Cons
  • Requires a multi-agent setup to be useful; not a standalone tool
  • The prompt is a template; users must configure their own sub-agents
  • No built-in execution; relies on external agent infrastructure

Best For

Complex task decomposition and delegationMulti-step research and implementation workflowsCode generation with review and testing pipelineData analysis and report writingAny multi-agent AI system requiring a central coordinator

FAQ

What is the Multi-Agent Orchestrator prompt?
It is a system prompt for an AI agent that acts as a central dispatcher, decomposing tasks and delegating to specialized sub-agents.
How does the orchestrator handle dependencies between sub-tasks?
It identifies dependencies and orders execution: parallel for independent tasks, sequential for dependent ones, with a chaining pattern to pass outputs.
What sub-agents are available?
The prompt defines default agents: researcher, coder, reviewer, data_analyst, writer. Users can extend or replace them.
Can the orchestrator execute tasks directly?
No, it is strictly a router and coordinator; it never executes tasks directly.