Keboola MCP Server

Example Use Cases

• 1

  Query and manage Keboola storage buckets and tables directly from AI assistants.

• 2

  Create, run, and monitor SQL transformations, components, jobs, and orchestrator flows programmatically.

• 3

  Search, read, and update project metadata and deploy Streamlit data apps to present query results.

Description

Keboola MCP Server turns Keboola platform features (storage, components, SQL, jobs, flows, data apps, and metadata) into tools that AI assistants like Claude, Cursor, and others can call. It supports both a hosted remote MCP Server and local development modes with multiple transport options (stdio, streamable-http, http-compat). The server respects Keboola development branches and can be run via Docker, integrated client launches, or manual CLI for development and testing. This lets agents query tables, create and run transformations, manage components and workflows, and surface project documentation programmatically.