mcp-server-datahub

by acryldata

Data/Management metadata datahub ai agents data ecosystem

An implementation of a Model Context Protocol (MCP) server for DataHub, enabling AI agents to query DataHub for metadata and context about a data ecosystem. It supports both DataHub Core and DataHub Cloud.

View on GitHub

Last updated: N/A

What is mcp-server-datahub?

mcp-server-datahub is a Model Context Protocol (MCP) server implementation designed to allow AI agents to interact with and query DataHub for metadata and context about your data ecosystem.

How to use mcp-server-datahub?

To use mcp-server-datahub, install uv, locate your DataHub instance URL and personal access token, and configure your MCP client (e.g., Claude Desktop, Cursor) with the provided configuration snippets, ensuring the correct paths and environment variables are set.

Key features of mcp-server-datahub

Searching across all entity types and using arbitrary filters
Fetching metadata for any entity
Traversing the lineage graph, both upstream and downstream
Listing SQL queries associated with a dataset

Use cases of mcp-server-datahub

Enabling AI agents to understand and interact with data assets in DataHub
Providing context to AI agents for data-related tasks
Automating metadata discovery and analysis
Integrating DataHub metadata into AI-powered workflows

FAQ from mcp-server-datahub

What is uv?

uv is a fast, modern Python package installer and resolver.

Where do I find my DataHub URL?

Your DataHub URL is the address of your DataHub instance, e.g., https://tenant.acryl.io/gms.

How do I get a DataHub personal access token?

You can create a personal access token in DataHub following the instructions in the DataHub documentation.

What if I get a `spawn uvx ENOENT` error?

This means the uvx command is not found. Replace uvx with the full path to the uvx executable, which you can find using which uvx.

Can I use this with DataHub Core?

Yes, mcp-server-datahub supports both DataHub Core and DataHub Cloud.