ChromaDB MCP Server

Example Use Cases

• 1

  Retrieve semantically similar document chunks for question-answering or retrieval-augmented generation.

• 2

  Integrate a local MCP tool into Claude Desktop or other MCP-capable clients to enable document lookups.

• 3

  Build a searchable vector database over a directory of files using Google embeddings and ChromaDB.

Description

This project ingests documents from a directory, chunks them, generates semantic embeddings with Google's text-embedding-004 model, and stores them in a persistent ChromaDB vector database. It exposes an MCP tool to query for similar chunks (get_similar_chunks) so other agents or tools can perform retrieval-augmented tasks. The server is designed for easy integration (e.g., with Claude Desktop) and simplifies building semantic search capabilities over local document collections.