Tathagata003/mcp_servers

csv-parquet MCP Server Tools

This project implements several MCP server tools for working with CSV, Parquet, PostgreSQL, and AWS S3 data. It is designed to be used as a backend for Claude Desktop, enabling natural language access to your data via MCP.

Features

• CSV & Parquet Summaries:
  Quickly get row/column counts for files in the data/ directory.
• PostgreSQL Table Tools:
  • Table summary statistics
  • Filter rows by column value
  • List table columns
  • Get unique values from a column
• AWS S3 Tools:
  • List buckets and files
  • Download/upload files

Usage

1. Install Dependencies

cd D:/mcp/csv_parquet
python -m pip install -r requirements.txt
# Or use uv for faster installs:
uv pip install -r requirements.txt

2. Configure Environment

Edit .env with your database and AWS credentials.

3. Run the MCP Server

You can run the server with either Python or uv:

With Python

python [server.py](http://_vscodecontentref_/0) 

4. Connect with the Claude Desktop

Create the following file at C:\Users\%username%\AppData\Roaming\Claude\desktop_claude_config.json:

{
  "mcpServers": 
  {
    "mix_server": 
    {
      "command": "D:/mcp/csv_parquet/env/Scripts/python.exe",
      "args": [
        "D:/mcp/csv_parquet/server.py"
      ]
    }
  }
}

Claude Desktop will now detect and connect to your MCP server, allowing you to use natural language queries for your data.