🗂️ File Manager MCP

Open-Source MCP Server for Local File Management

File Manager MCP is an open-source Model Context Protocol (MCP) server that enables AI assistants such as ChatGPT, Cursor, VS Code, Windsurf, and other MCP-compatible clients to perform local filesystem operations through natural language.

Built with FastMCP and Streamable HTTP transport, the server exposes 35 filesystem tools covering file management, directory operations, search, ZIP handling, batch processing, and utility operations.

🚀 Why File Manager MCP?

Instead of manually navigating files and folders, users can interact with their filesystem through AI.

Example

User:

Create a FastAPI project structure

AI:

✓ Creates folders
✓ Creates README.md
✓ Creates requirements.txt
✓ Creates source files

User:

Zip the project and move it to backups

AI:

✓ Creates ZIP archive
✓ Moves archive

Everything runs locally on the user's machine.

Related MCP server: folder-mcp

✨ Features

File Operations

• Create files

• Read files

• Update files

• Append content

• Delete files

• Move files

• Copy files

• Rename files

Directory Operations

• Create directories

• Delete directories

• Analyze directories

• Generate directory trees

Search Services

• Recursive file search

• Folder search

• Metadata inspection

ZIP Operations

• Create ZIP archives

• Extract ZIP archives

• List ZIP contents

Batch Operations

• Batch create files

• Batch delete files

• Batch move files

• Batch copy files

• Batch rename files

• Batch create directories

• Batch delete directories

Utility Operations

• File hashing

• Duplicate detection

• File comparison

• Directory comparison

• Large file discovery

• Empty directory cleanup

• Temporary file cleanup

📊 Project Statistics

🏗️ Architecture

User
 │
 ▼
MCP Client
(ChatGPT / Cursor / VS Code)
 │
 ▼
File Manager MCP
(FastMCP Server)
 │
 ├── File Services
 ├── Search Services
 ├── Batch Services
 ├── ZIP Services
 └── Utility Services
 │
 ▼
Local Filesystem

🛠️ Technology Stack

• Python

• FastMCP

• Model Context Protocol (MCP)

• Streamable HTTP Transport

• pathlib

• shutil

• zipfile

• hashlib

• logging

🚀 Quick Start

Clone Repository

git clone https://github.com/parthRana26/file-manager-mcp.git

cd file-manager-mcp

Install uv

pip install uv

Install Dependencies

uv sync

Start Server

uv run main.py

Server Endpoint:

http://localhost:8000/mcp

🔍 Testing with MCP Inspector

Start MCP Inspector:

npx @modelcontextprotocol/inspector

Connect using:

Transport: Streamable HTTP

URL:
http://localhost:8000/mcp

Verify that all tools appear successfully.

🔌 Connecting MCP Clients

Use the following endpoint:

http://localhost:8000/mcp

Example configuration:

{
  "mcpServers": {
    "file-manager-mcp": {
      "type": "http",
      "url": "http://localhost:8000/mcp"
    }
  }
}

Works with:

• ChatGPT

• Cursor

• VS Code

• Windsurf

• MCP Inspector

• Custom MCP Clients

💬 Example Prompts

Create a README.md file

Generate a FastAPI project structure

Find duplicate files

Create a ZIP archive of my project

Compare two directories

Find files larger than 500 MB

📁 Project Structure

file-manager-mcp/

├── main.py
├── requirements.txt
├── config.py
├── README.md

├── services/
│   ├── file_service.py
│   ├── search_service.py
│   ├── batch_service.py
│   ├── zip_service.py
│   ├── utility_service.py
│   └── security_service.py

└── workspace/

🔒 Security Notes

This MCP server performs real filesystem operations.

Recommended:

• Run locally

• Use trusted MCP clients

• Review destructive operations before approval

• Use project-specific directories

👨‍💻 Maintainer

Parth Rana

GitHub: https://github.com/parthRana26

LinkedIn: https://linkedin.com/in/parth-rana-profile

🤝 Contributing

Contributions, issues, and feature requests are welcome.

Feel free to fork the repository and submit pull requests.

📄 License

Licensed under the MIT License.

See LICENSE for details.