ibm-watsonxdata-mcp-server

Overview

The IBM watsonx.data MCP Server provides a bridge between AI assistants and IBM watsonx.data lakehouses. It organizes lakehouse operations into six primary categories: platform management, engine operations, catalog management, query execution, Spark application handling, and data ingestion. By utilizing IBM Cloud IAM authentication and supporting transports like stdio and streamable HTTP, it lets users run complex queries, manage Presto and Spark engines, inspect schemas, and load data conversationally.

Capabilities

• ▸Platform management tools for instance status and configuration
• ▸Engine operations for Presto and Spark lifecycle management
• ▸Catalog tools for schema/table discovery and DDL execution
• ▸Query execution tools supporting SELECT, INSERT, UPDATE, and query plan analysis
• ▸Spark application submission and monitoring
• ▸Data ingestion tools for loading CSV, Parquet, and JSON from object storage
• ▸IBM Cloud IAM token-based authentication with automatic refresh

Best for

Checking instance status and metadata details, Managing and monitoring Presto and Spark compute engines, Browsing database catalogs, schemas, and table structures, Executing SQL queries and analyzing query plans, Submitting and monitoring Apache Spark jobs, Ingesting data from object storage into lakehouse tables

Works with

Security considerations

Relies on IBM Cloud API keys and IAM tokens. The streamable HTTP transport does not include a built-in authentication layer, meaning binding to non-localhost interfaces requires an external reverse proxy with auth to prevent unauthorized access.

Categories