Vectorize MCP Server

This repository contains an implementation of a Model Context Protocol (MCP) server designed to integrate with Vectorize for vector retrieval and text extraction capabilities.

Installation

Running with npx

To run the server, set the environment variables VECTORIZE_ORG_ID, VECTORIZE_TOKEN, and VECTORIZE_PIPELINE_ID with your Vectorize credentials and pipeline ID. Then, execute the provided npx command to start the server.

Configuration

The provided JSON configuration snippet demonstrates how to configure the server within environments like Claude/Windsurf/Cursor/Cline. It specifies the command to execute (using npx), arguments, and environment variables required for the Vectorize integration.

Tools

The server offers three primary tools:

• Retrieve documents: Performs vector search using a question and retrieves the top k relevant documents.
• Text extraction: Extracts text from documents (e.g., PDFs) and chunks it into Markdown format.
• Deep Research: Generates a private deep research report from your pipeline based on a query.

Development

To contribute:

1. Fork the repository.
2. Create a feature branch.
3. Submit a pull request.