Happy slice of mcp pizzaMCP.pizza
Fire in da houseTop Tip:Paying $100+ per month for Perplexity, MidJourney, Runway, ChatGPT is crazy - get all your AI tools in one site starting at $15 per month with Galaxy AIFire in da houseCheck it out free

manusmcp

MCP.Pizza Chef: mantrakp04

ManusMCP is an MCP server framework that orchestrates specialized AI agents with distinct roles like Planner, FileWizard, CommandRunner, and WebNavigator. It uses Flowise to deploy these agents collaboratively, enabling complex problem-solving and task automation through shared context and communication. ManusMCP supports multi-agent workflows and is designed for seamless integration with LLMs, tested primarily with Anthropic models.

Use This MCP server To

Coordinate multiple AI agents for complex task workflows Automate multi-step problem-solving with specialized AI roles Deploy AI agents for file management and command execution Enable AI-driven web navigation and data extraction Integrate AI agent orchestration into custom applications Manage AI workflows using Flowise UI for configuration Run AI agent teams for task planning and execution

README

ManusMCP

An AI agent framework using Flowise to deploy AI team members with specialized capabilities.

Note: Tested with Anthropic models only. Reason

What is ManusMCP?

ManusMCP is a framework for orchestrating specialized AI agents that work together to accomplish complex tasks. The Model Context Protocol (MCP) enables these agents to collaborate effectively by providing a shared context and communication layer.

Installation

# Clone the repository
git clone https://github.com/mantrakp04/manusmcp.git
cd manusmcp/pytools

# Install dependencies
npm i -g bun
bash flowise.sh

Usage

1: Run it with docker

cp .env.example .env
docker-compose up -d

2: Open Flowise

  1. Open Flowise UI in your browser
  2. Go to "Agentflows" section
  3. Click on "Add New" (Top Right)
  4. Click Settings (Top Right)
  5. Click Import and select the v2.2 and save it
  6. Create a new flow, (FROM STEP 2)
  7. Use v2.3 as the flow file

[Optional] Checkout the MCP Server

cd pytools
bunx @modelcontextprotocol/inspector uv run manusmcp

Recommended runtime: bun

Contributing

This project is currently in development with initial working functionality. Here are the core components that could benefit from community contributions:

  • Prompts: Help refine the AI agent prompts for more effective interactions
  • Workflows: Enhance the workflow definitions in Agents.json
  • MCP Server: Improve the Model Context Protocol server implementation in .runtime/index.ts

To contribute:

  1. Fork the repository
  2. Create a feature branch
  3. Submit a pull request with your changes

Sources

  • Huggingface discord post

Disclaimer

Currently testing in main branch.

manusmcp FAQ

How do I install ManusMCP?
Clone the repository, install dependencies with npm and bun, then run the docker-compose setup as documented.
What AI models does ManusMCP support?
ManusMCP has been tested primarily with Anthropic models but can be adapted to work with OpenAI, Claude, and Gemini models.
How do I configure AI agent workflows in ManusMCP?
Use the Flowise UI to import agent flows, configure settings, and manage specialized AI agents.
Can ManusMCP handle multiple AI agents simultaneously?
Yes, ManusMCP orchestrates multiple specialized AI agents working collaboratively on complex tasks.
Is ManusMCP limited to specific AI agent roles?
No, it currently includes roles like Planner, FileWizard, CommandRunner, and WebNavigator but can be extended with custom agents.
How does ManusMCP facilitate communication between AI agents?
It uses the Model Context Protocol (MCP) to provide a shared context and communication layer for agent collaboration.
Can I run ManusMCP without Docker?
While Docker is recommended for ease, you can run ManusMCP by manually installing dependencies and running the services.
What is Flowise in the context of ManusMCP?
Flowise is the UI and orchestration tool used to design, deploy, and manage AI agent workflows within ManusMCP.