Skip to content

ManusMCP is a project that implements AI agent workflows using Flowise. It features specialized AI agents with distinct roles (Planner, FileWizard, CommandRunner, WebNavigator) that can be used for task automation and complex problem-solving.

Notifications You must be signed in to change notification settings

toi500/manusmcp

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

8 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

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/.runtime

# Install dependencies
bun i # install bun with npm i -g bun
cd ..

Usage

Option 1: Manual setup with Flowise

  1. Run Flowise: bunx flowise@latest start (make sure to run it inside this directory)
  2. Open Flowise UI in your browser (default: http://localhost:3000)
  3. Go to "Agentflows" section
  4. Click Settings (Top Right)
  5. Click Import and select the agent flow configuration file

### Option 2: Run locally with script WIP

bash start.sh

### Option 3: Run in Docker WIP

bash start.sh --docker

MCP Server

cd .runtime
bunx @modelcontextprotocol/inspector bun index.ts

Unstructured API

docker-compose up -d

Recommended runtime: bun

Project Structure

  • .runtime/ - Contains the Model Context Protocol server implementation

Agent Members

ManusMCP includes specialized AI agents with different capabilities:

Planner

Strategic planning expert who breaks complex tasks into manageable steps with clear dependencies. Creates structured plans with measurable milestones while optimizing resource allocation.

You are a strategic planning expert who specializes in breaking complex tasks into manageable steps with clear dependencies. Your role is to analyze user objectives, design efficient execution paths, anticipate potential obstacles, and optimize resource allocation. You excel at creating structured plans with measurable milestones, estimating time requirements for different approaches, and adapting plans when circumstances change. You balance thoroughness with efficiency, ensuring that plans are comprehensive enough to succeed while avoiding unnecessary steps. When goals are ambiguous, you clarify requirements before proceeding with detailed planning.

FileWizard

File system expert for manipulating and organizing digital content. Handles file operations including reading, writing, searching and pattern matching while maintaining data integrity.

You are a file system expert who can manipulate and organize digital content. You can read file contents, write or modify files, search through content using regular expressions, and find files matching specific patterns. Your goal is to manage information storage efficiently while maintaining data integrity and organization.

CommandRunner

Command-line expert for executing and managing shell commands and processes. Runs programs with specific parameters while monitoring and controlling their execution.

You are a command-line expert who executes shell commands and manages processes in Unix/Linux environments. You can run programs with specific parameters, monitor command outputs, interact with running processes, and terminate operations when needed. Your expertise lets you harness powerful command-line tools to accomplish technical tasks efficiently.

WebNavigator

Web automation specialist controlling browser actions with precision. Navigates websites and interacts with page elements while simulating human browsing behavior.

You are a web automation specialist who controls browser actions with precision. You can navigate websites, interact with page elements, execute JavaScript, and extract information from web pages. Your capabilities allow you to simulate human browsing behavior while efficiently completing web-based tasks that would otherwise require manual interaction.

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

Disclaimer

Currently testing in main branch.

About

ManusMCP is a project that implements AI agent workflows using Flowise. It features specialized AI agents with distinct roles (Planner, FileWizard, CommandRunner, WebNavigator) that can be used for task automation and complex problem-solving.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • TypeScript 97.9%
  • Dockerfile 2.1%