DataHub MCP Server is a Model Context Protocol server implementation that sits atop DataHub’s metadata platform to empower AI agents and developers with semantic data discovery and provenance capabilities. It exposes a set of MCP tools that enable searching across datasets, assets, and metadata using structured queries, retrieving lineage paths, and inspecting schema fields. With features like Structured Search with Context Filtering, users can use wildcard, field-specific searches, and Boolean logic to narrow results. The SQL intelligence component supports accessing existing queries and generating new ones, helpful for understanding join patterns, filters, and aggregation behavior. The server also exposes table- and column-level lineage to trace data flow, transformations, and dependencies across upstream and downstream nodes (including multiple hops). It additionally provides a view into the data ecosystem by exposing domains, owners, tags, glossaries, and data platforms, so teams can understand context before searching. The documentation references usage in the DataHub MCP server docs and demonstrates an end-to-end agent workflow that searches, inspects metadata, inspects lineage, retrieves example queries, and constructs SQL.