An MCP server that provides vector database capabilities through Chroma, enabling semantic document search, metadata filtering, and document management with persistent storage.

Installation

UV Package Manager

uv venv
uv sync --dev --all-extras

Running the Server

uv run chroma

Configuration

Claude Desktop

{
  "mcpServers": {
    "chroma": {
      "command": "uv",
      "args": [
        "--directory",
        "C:/MCP/server/community/chroma",
        "run",
        "chroma"
      ]
    }
  }
}

Available Tools

Tool	Description
create_document	Create a new document with content and optional metadata
read_document	Retrieve a document by ID
update_document	Update the content and metadata of an existing document
delete_document	Delete a document from the database
list_documents	List all documents with optional limit and offset
search_similar	Find semantically similar documents using vector search

Features

• Semantic search using Chroma embeddings
• Metadata filtering for targeted search
• Content filtering for additional document filtering
• Persistent storage in local directory
• Comprehensive error handling with clear messages
• Retry logic for temporary failures

Usage Examples

Create a document about machine learning with metadata fields such as year and topic

Search for documents similar to 'machine learning models' with metadata filters

Find semantically similar documents based on content meaning

Store and retrieve documents with custom metadata for organization

Notes

Data is stored in the 'src/chroma/data' directory. The server can be tested using MCP Inspector. Requires Python 3.8+, Chroma 0.4.0+, and MCP SDK 0.1.0+.