MCP Server: Local Ai Tooling With Python

🚀 🧠 MCP Local AI Tool Server: Implemented in Python

This project sets up a lightweight Model Context Protocol (MCP) server using Python, allowing AI assistants like Claude to interact with local data files (such as CSV and Parquet). By integrating with the desktop version of Claude, you can perform tasks like summarizing datasets or retrieving specific information through natural language queries.

🚀 Quick Start

This project provides a local MCP server that enables seamless interaction between AI assistants and local data files. After installation and configuration, you can use natural - language prompts to interact with data through Claude.

✨ Features

• Local MCP Server (mix_server): Enables communication between Claude and local tools.
• CSV and Parquet Data Readers: Used to load and process files in common data analysis tabular formats.
• Modular Structure: The codebase is designed simply, making it easy to add more tools.
• Smooth Integration with Claude: Connects directly to the desktop version of Claude, allowing data interaction through natural language.

📦 Installation

1. Clone the Repository

git clone https://github.com/yourusername/mcp-server.git
cd mcp-server

2. Set Up the Environment with uv

uv is a fast Python package and project management tool written in Rust. As a replacement for pip, pip - tools, and virtualenv, it offers significant speed improvements and a unified interface.

Install uv

You can install uv using any of the following methods:

• Using curl (macOS and Linux):

  curl -LsSf https://astral.sh/uv/install.sh | sh

• Using pip:

  pip install uv

• Using pipx:

  pipx install uv

For more installation options and detailed information, refer to the official uv documentation.

Create a Virtual Environment and Install Dependencies

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv pip install mcp[cli] pandas pyarrow

This will create a virtual environment and install the necessary dependencies in it using uv.

📁 Project Structure

mcp-server/
├── tools/
│   ├── csv_reader.py
│   └── parquet_reader.py
├── mix_server.py
├── README.md
└── requirements.txt

• tools/: Contains independent tool scripts for data processing.
• mix_server.py: Initializes and runs the MCP server, registering available tools.

⚙️ Desktop - Version Claude Configuration

To integrate with the desktop version of Claude, add the following configuration to your Claude settings:

{
  "mcpServers": {
    "mix_server": {
      "command": "/absolute/path/to/.venv/bin/python",
      "args": ["/absolute/path/to/mix_server.py"]
    }
  }
}

Make sure the paths to the Python interpreter and mix_server.py are correct.

💻 Usage Examples

Basic Usage

After setting up the project, you can use natural - language prompts to interact with data through Claude. For example:

• "Summarize the contents of data.csv."
• "How many rows and columns are there in dataset.parquet?"

Claude will use the registered tools to process your requests and provide the required information.

📚 Documentation

This project is based on the following tutorial: [Building a Basic MCP Server with Python](https://medium.com/data - engineering - with - dremio/building - a - basic - mcp - server - with - python - 4c34c4101ed).

📄 License

This project is licensed under the MIT License.