Local MCP Server

Example Use Cases

• 1

  Summarize the contents of local CSV or Parquet datasets and return human-readable overviews.

• 2

  Report basic dataset metadata (row/column counts, column types) from files on the user’s machine.

• 3

  Extract or filter rows from local tabular files in response to natural language queries.

Description

This project provides a small Model Context Protocol (MCP) server written in Python that exposes local tools for reading CSV and Parquet datasets. It registers simple reader tools so Claude for Desktop (or other MCP-capable agents) can invoke them to summarize, inspect, or extract data via natural language. The design is modular and easy to extend with additional tools, enabling fast local data exploration without uploading sensitive files to external services.