Federa3911/multi-agent-orchestrator
Orchestrate multi-agent workflows with LangGraph for Python 3.11+, with a lightweight, tested framework for routing tasks and agents
Platform-specific configuration:
{
"mcpServers": {
"multi-agent-orchestrator": {
"command": "npx",
"args": [
"-y",
"multi-agent-orchestrator"
]
}
}
}Add the config above to .claude/settings.json under the mcpServers key.
[](https://github.com/Federa3911/multi-agent-orchestrator)
multi-agent-orchestrator helps you run a set of AI agents from one place. It uses a supervisor pattern, which means one main agent helps guide the others. This setup is useful when you want different agents to handle different tasks, like research, writing, planning, or web lookup.
The app is built for Windows users who want to get started fast. You can download the project, open it, and run it with a simple local setup. It uses LangGraph under the hood, and it can work with tools like OpenAI, Anthropic, and Tavily.
Before you start, make sure your PC has:
If you plan to use the app with larger models or more agents, 16 GB of RAM gives a smoother experience.
Visit this page to download and use the project:
https://github.com/Federa3911/multi-agent-orchestrator
Follow these steps on your Windows PC:
If the project uses a virtual environment, it keeps the app files separate from the rest of your PC. That helps avoid conflicts with other Python apps.
Use these steps after you unpack the files:
Loading reviews...