proj_mcp_synthetic_data

syahriikram/proj_mcp_synthetic_data

3.1

If you are the rightful owner of proj_mcp_synthetic_data and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

This document provides a comprehensive overview of setting up a Model Context Protocol (MCP) server for synthetic data generation using the SDV library.

Tools
3
Resources
0
Prompts
0

mcp_sdg

Creating MCP server for synthetic data generation using SDV library.

Setup

Install uv as the main package manager

brew install uv
uv init
uv venv
source .venv/bin/activate

Packages

uv add sdv fastmcp pandas

Run inspector to test locally: npx @modelcontextprotocol/inspector uv run main.py

Install Claude Desktop as MCP Client (or any other MCP Clients).

  • Update claude_desktop_config.json for Claude Desktop to connect to the MCP Server (locally hosted)

MCP Tools

  1. Generate Tool: creates synthetic data from real data using SDV Synthesizer
  2. Evaluate Tool: evaluates quality of synthetic data in comparison to real data by assessing statistical similarity to determine which real data patterns are captured by the synthetic data
  3. Visualize Tool: generates visualization to compare real and synthetic data for specific column

Execution

Prompts to run in Claude Desktop:

  1. Generate synthetic data for the data present in the folder "..."
  2. Evaluate the synthetic data that has been generated for the actual data folder located at "..."
  3. Visualize amenities_fee column in the guests table, and compare the distribution of synthetic data to that of real data for this specific column

Technology Stack

Python

Core Concepts

AI, MCP

Can you help me do the same?

Happy to work with you, contact me at syahriikram@gmail.com

GitHub Syahri