IBM watsonx.data MCP Server is a Model Context Protocol server that enables AI assistants to query and explore IBM watsonx.data lakehouses using natural language.
The tool solves the problem of integrating lakehouse data access into AI assistant workflows. It provides a set of specialized tools organized into six functional categories: platform management for instance status and configuration, engine operations for managing Presto and Spark engines, catalog management for browsing schemas and tables, query execution for running SELECT, INSERT, and UPDATE queries with query plan analysis, Spark application management for submitting and monitoring jobs, and data ingestion for loading data from object storage. The server uses the Model Context Protocol to expose these capabilities to AI assistants, handling IBM Cloud IAM authentication with automatic token refresh and supporting both stdio transport for local subprocess and streamable HTTP transport.
Developers should adopt this tool if they need to give AI assistants direct access to watsonx.data lakehouses without building custom integrations. It suits projects where natural language querying of lakehouse data is valuable and where the deployment environment can support either stdio or HTTP transport. The README notes that IBM watsonx.data also provides a hosted remote MCP server that requires no installation, offering an alternative for users who prefer not to run their own server instance. The tool is compatible with Claude Desktop, IBM Bob, and other MCP-enabled AI assistants.
Development activity shows consistent engagement with the codebase through regular updates and maintenance of the tool's core functionality. The project maintains active documentation including a dedicated TOOLS.md file detailing all available capabilities and their usage patterns. Security and authentication mechanisms receive ongoing attention, with IBM Cloud IAM integration and token refresh handling built into the implementation. The project provides multiple installation paths including both pip and development setup options, indicating attention to different user deployment scenarios.
