Agency Orchestrator

v1.0.0Coding Agentsstable

🚀 One sentence → multi-AI-role collaboration → complete plan in minutes. 211+ expert roles, zero-code YAML, 9 LLM providers (6 free). 一句话调度多个AI专家自动协作,几分钟交付完整方案。

agent-orchestrationai-agentsaiderautogen-alternativeclaude
Share:
980
Stars
0
Downloads
0
Weekly
0/5

What is Agency Orchestrator?

Agency Orchestrator is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to 🚀 one sentence → multi-ai-role collaboration → complete plan in minutes. 211+ expert roles, zero-code yaml, 9 llm providers (6 free). 一句话调度多个ai专家自动协作,几分钟交付完整方案。

🚀 One sentence → multi-AI-role collaboration → complete plan in minutes. 211+ expert roles, zero-code YAML, 9 LLM providers (6 free). 一句话调度多个AI专家自动协作,几分钟交付完整方案。

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • 🚀 One sentence → multi-AI-role collaboration → complete plan

Use Cases

Multi-expert role collaboration from text prompt
Zero-code YAML with 9 LLM provider support
jnMetaCode

Maintainer

LicenseApache-2.0
Languagetypescript
Versionv1.0.0
UpdatedMay 22, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx agency-orchestrator

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Agency Orchestrator

Agency Orchestrator is a multi-agent AI collaboration framework that turns a single natural language prompt into a coordinated workflow involving up to 211 specialized expert roles — from architect and security reviewer to UX designer and DevOps engineer. It uses YAML-based DAG workflows for dependency management and parallel execution, supports 10 LLM providers (including 7 that require no API keys), and ships with 32 built-in workflow templates for software development, marketing, operations, and strategy. Developers and product teams use it to automate complex multi-step deliverables that would otherwise require chaining many AI conversations manually.

Prerequisites

  • Node.js 18+ with npm for installation
  • At least one LLM provider: Claude Code, Gemini CLI, Ollama (free/no key), or API keys for DeepSeek (DEEPSEEK_API_KEY), Claude (ANTHROPIC_API_KEY), or OpenAI (OPENAI_API_KEY)
  • An MCP client such as Claude Desktop or Cursor for MCP integration
  • Optional: YAML editing capability for creating custom workflows
1

Install Agency Orchestrator globally

Install the agency-orchestrator CLI globally via npm to get the 'ao' command available system-wide.

npm install -g agency-orchestrator
2

Configure your LLM provider

Set environment variables for your preferred LLM provider. DeepSeek is recommended for cost-performance balance. Free CLI-based providers like claude-code or gemini-cli require no API key.

# For DeepSeek (recommended paid option)
export DEEPSEEK_API_KEY=your_key_here

# For Claude API
export ANTHROPIC_API_KEY=your_key_here

# For OpenAI or compatible providers
export OPENAI_API_KEY=your_key_here
export OPENAI_BASE_URL=https://your-endpoint.com  # optional
3

Initialize a project

Run 'ao init' in your project directory to create a configuration file and select your default LLM provider interactively.

ao init
4

Run a workflow from a natural language prompt

Use 'ao compose' to have the orchestrator automatically design and run a multi-agent workflow from your description.

ao compose "analyze the feasibility of building an AI record-keeping tool" --run
5

Configure as an MCP server

Add Agency Orchestrator to your MCP client config to invoke multi-agent workflows as MCP tool calls from within Claude Desktop or Cursor.

{
  "mcpServers": {
    "agency-orchestrator": {
      "command": "npx",
      "args": ["agency-orchestrator"],
      "env": {
        "DEEPSEEK_API_KEY": "your_key_here"
      }
    }
  }
}
6

Use built-in workflow templates

Browse and run the 32 built-in templates for common tasks like PR reviews, marketing plans, and architecture designs.

# List available templates
ao list

# Run a PR review workflow
ao run workflows/dev/pr-review.yaml --input code=@src/main.ts

Agency Orchestrator Examples

Client configuration

MCP configuration for Claude Desktop to use Agency Orchestrator with DeepSeek as the LLM provider.

{
  "mcpServers": {
    "agency-orchestrator": {
      "command": "npx",
      "args": ["agency-orchestrator"],
      "env": {
        "DEEPSEEK_API_KEY": "your_deepseek_key_here"
      }
    }
  }
}

Prompts to try

Example prompts and CLI commands demonstrating multi-agent orchestration capabilities.

- "Compose a complete go-to-market strategy for a developer productivity SaaS tool"
- "Review the security and architecture of this API design using expert roles"
- Run: ao compose "write and test a Python web scraper" --run
- Run: ao run workflows/dev/pr-review.yaml --input code=@src/api.ts
- Run: ao compose "create a content calendar for Q3" --run --from marketing_strategy

Troubleshooting Agency Orchestrator

LLM provider authentication fails

Verify the correct environment variable is set for your provider: DEEPSEEK_API_KEY, ANTHROPIC_API_KEY, or OPENAI_API_KEY. Run 'ao init' again to reconfigure the provider selection if needed.

Workflow hangs waiting for a specific agent role

Use 'ao run workflow.yaml --resume last' to restart from the last checkpoint. Check the DAG for cyclic dependencies in your custom YAML if building your own workflows.

'ao' command not found after installation

Ensure npm global bin directory is in your PATH. Run 'npm config get prefix' to find the global bin location, then add '<prefix>/bin' to your PATH in ~/.bashrc or ~/.zshrc.

Frequently Asked Questions about Agency Orchestrator

What is Agency Orchestrator?

Agency Orchestrator is a Model Context Protocol (MCP) server that 🚀 one sentence → multi-ai-role collaboration → complete plan in minutes. 211+ expert roles, zero-code yaml, 9 llm providers (6 free). 一句话调度多个ai专家自动协作,几分钟交付完整方案。 It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Agency Orchestrator?

Follow the installation instructions on the Agency Orchestrator GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Agency Orchestrator?

Agency Orchestrator works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Agency Orchestrator free to use?

Yes, Agency Orchestrator is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "agency-orchestrator": { "command": "npx", "args": ["-y", "agency-orchestrator"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Agency Orchestrator?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides