Skip to main content
Watch: I Built an AI App Without Leaving Databricks

Data and BI - official IBM server

watsonx.data Document Library Retrieval MCP server

MCP server that discovers watsonx.data document libraries, registers each as an MCP tool and answers natural language queries against them. It runs locally from ibm-watsonxdata-dl-retrieval-mcp-server (pypi) with API key auth.

Last verified 3 sources

Install

Generic

Terminal
pip install ibm-watsonxdata-dl-retrieval-mcp-server

Notes and caveats

  • Runs in SSE mode on port 8000 by default, or over stdio with --transport stdio.
  • Watsonx SaaS uses an IBM Cloud API key; Cloud Pak for Data uses CPD username, password and a CA bundle.
  • Each document library is registered as its own tool named <library_name><library_id>.
  • IBM's catalog also lists a remote version hosted in IBM infrastructure.

FAQ

What does the watsonx.data Document Library Retrieval MCP server do?

MCP server that discovers watsonx.data document libraries, registers each as an MCP tool and answers natural language queries against them.

How do I add the watsonx.data Document Library Retrieval MCP server to an MCP client?

Use pip install ibm-watsonxdata-dl-retrieval-mcp-server, copied from IBM's docs.

Does the watsonx.data Document Library Retrieval MCP server need authentication?

Yes, API key. Transport: stdio, SSE. Status: Unlabelled, last checked October 17, 2026.

Sources

Last verified against 3 IBM pages, then re-checked independently.

Newsletter

New MCP servers, the week they ship

One email per week. New official MCP servers, the coding tools that support them and how developers wire them up.