juanmaalt/csv_analyzer_mcp_server

CSV Analyzer MCP Server

A lightweight, modular tool designed to analyze CSV datasets and produce structured summaries suitable for consumption by language models (LLMs) like ChatGPT or Claude.

Features

• CSV Content Analysis: Processes CSV data provided as raw content.
• Data Cleaning: Options to remove duplicate entries and rows with missing values.
• Flexible Output: Returns results in JSON or Markdown format.
• Customizable Delimiters: Supports various CSV delimiters.

Installation

1. Clone the repository:

   git clone https://github.com/juanmaalt/csv_analyzer_mcp_server.git
   cd csv_analyzer_mcp_server
   

2. Install dependencies using uv:

   uv pip install -r requirements.txt
   

   Note: Ensure you have uv installed. If not, refer to the uv installation guide.

Usage

The primary function analyze_csv can be invoked with the following parameters:

• csv_content (str): Raw CSV data as a string.
• delimiter (str): Delimiter used in the CSV (default: ,).
• remove_duplicates (bool): Remove duplicate rows (default: True).
• remove_non_valid_data (bool): Remove rows with missing values (default: True).
• output_format (str): Output format - 'json' or 'markdown' (default: 'json').

Project Structure

csv_analyzer_mcp_server/
├── core/               # Core functionality
├── data/               # Sample CSV files
├── main.py             # Entry point
├── pyproject.toml      # Project metadata
├── uv.lock             # Dependency lock file for uv
└── .gitignore          # Git ignore rules

Contributing

Contributions are welcome! Please fork the repository and submit a pull request.