mcp-server-datahub

A Model Context Protocol server implementation for DataHub. This enables AI agents to query DataHub for metadata and context about your data ecosystem.

Supports both DataHub Core and DataHub Cloud.

• 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

Check out the demo video, done in collaboration with the team at Block.

1. Install uv

   # On macOS and Linux.
   curl -LsSf https://astral.sh/uv/install.sh | sh

2. Locate your authentication details

   For authentication, you'll need the following:

   • The URL of your DataHub instance e.g. https://tenant.acryl.io/gms
   • A personal access token
   Alternative: Using ~/.datahubenv for authentication

   You can also use a ~/.datahubenv file to configure your authentication. The easiest way to create this file is to run datahub init and follow the prompts.

   uvx --from acryl-datahub datahub init

3. Configure your MCP client. See below - this will vary depending on your agent.

Run which uvx to find the full path to the uvx command.

In your claude_desktop_config.json file, add the following:

{
  "mcpServers": {
    "datahub": {
      "command": "<full-path-to-uvx>",  // e.g. /Users/hsheth/.local/bin/uvx
      "args": ["mcp-server-datahub"],
      "env": {
        "DATAHUB_GMS_URL": "<your-datahub-url>",
        "DATAHUB_GMS_TOKEN": "<your-datahub-token>"
      }
    }
  }
}

In .cursor/mcp.json, add the following:

{
  "mcpServers": {
    "datahub": {
      "command": "uvx",
      "args": ["mcp-server-datahub"],
      "env": {
        "DATAHUB_GMS_URL": "<your-datahub-url>",
        "DATAHUB_GMS_TOKEN": "<your-datahub-token>"
      }
    }
  }
}

command: uvx
args:
  - mcp-server-datahub
env:
  DATAHUB_GMS_URL: <your-datahub-url>
  DATAHUB_GMS_TOKEN: <your-datahub-token>

The full stack trace might look like this:

2025-04-08T19:58:16.593Z [datahub] [error] spawn uvx ENOENT {"stack":"Error: spawn uvx ENOENT\n    at ChildProcess._handle.onexit (node:internal/child_process:285:19)\n    at onErrorNT (node:internal/child_process:483:16)\n    at process.processTicksAndRejections (node:internal/process/task_queues:82:21)"}

Solution: Replace the uvx bit of the command with the output of which uvx.

See DEVELOPING.md.