mcp-rag-search-server

mcp-rag-search-server

2
research_and_datadeveloper_tools

This project is a custom Model Calling Protocol (MCP) server providing RAG capabilities and multiple web search tools. It integrates Google's Gemini 2.0 for AI-powered searches and Linkup for traditional search methods.

MCP Server with RAG and Multi-Search

A custom MCP server offering RAG capabilities and multiple web search options through Google's Gemini 2.0 API and Linkup.

Features

  • RAG workflow using local documents
  • Advanced web search with Google's Gemini 2.0 and traditional search with Linkup

Prerequisites

  • Python 3.8+
  • Ollama and DeepSeek models
  • Gemini API key

Setup

  • Clone repository and install dependencies
  • Configure API keys in .env file

Usage

  • web_search for best available search
  • gemini_search for Gemini 2.0
  • linkup_search for Linkup
  • rag for local documents search