LanceDB MCP Server

Example Use Cases

• 1

  Ingest large sets of documents, compute and store embeddings, and make them available for later semantic retrieval using LanceDB.

• 2

  Perform vector-based document retrieval to answer user queries or provide contextual information from a stored corpus.

• 3

  Inspect or manage LanceDB tables (schema and metadata) to verify ingestion status and table details.

Description

This repository provides a minimal reference MCP server that integrates LanceDB to store embedded documents and serve retrieval requests. It exposes three primary capabilities: ingesting documents (with embeddings), retrieving relevant documents via vector search, and returning table details. It's intended as an example and starting point for building more complex MCP apps that rely on LanceDB-backed storage. Installation is performed by adding the supplied command configuration to a Claude MCP config.