MySearch-Proxy

Overview

MySearch-Proxy is a comprehensive search stack for AI assistants that consolidates multiple search providers into a single interface. It provides MCP tools for search, URL extraction, research workflows, and health monitoring. The server features a proxy layer for team deployment and token management, with support for both direct provider connections and a proxy-first architecture. The implementation includes advanced routing strategies, caching, and fallback mechanisms to ensure reliable search results across different query types.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• search

  Unified search across Tavily, Firecrawl, Exa, and Social/X

  Try:Search for latest AI newsTry:Find documentation about React hooksTry:Search for research papers on climate change
• extract_url

  Extract content from a webpage URL

  Try:Extract content from https://example.com/articleTry:Get the full text from this blog postTry:Summarize the content at this URL
• research

  Perform comprehensive research with search and content extraction

  Try:Research the latest developments in quantum computingTry:Investigate the benefits of remote work for tech companiesTry:Analyze market trends in renewable energy
• mysearch_health

  Check health status of search providers

  Try:Check the health of all search providersTry:Are Tavily and Firecrawl working properly?Try:Get provider status and error details

Comparable tools

Installation

cd /path/to/MySearch-Proxy
python3 -m venv venv
source venv/bin/activate

# Configure environment variables
echo 'MYSEARCH_PROXY_BASE_URL=https://your-mysearch-proxy.example.com' >> .env
echo 'MYSEARCH_PROXY_API_KEY=mysp-...' >> .env

./install.sh

mkdir -p mysearch-proxy-data
docker run -d \
  --name mysearch-proxy \
  --restart unless-stopped \
  -p 9874:9874 \
  -e ADMIN_PASSWORD=change-me \
  -v $(pwd)/mysearch-proxy-data:/app/data \
  skernelx/mysearch-proxy:latest

{
  "mcpServers": {
    "mysearch": {
      "command": "python",
      "args": ["-m", "mysearch.mcp_server"],
      "env": {
        "MYSEARCH_PROXY_BASE_URL": "https://your-mysearch-proxy.example.com",
        "MYSEARCH_PROXY_API_KEY": "mysp-..."
      }
    }
  }
}

FAQ

What search providers are supported?

Currently supports Tavily, Firecrawl, Exa, and Social/X with options for domain filtering and result prioritization.

How does routing work between providers?

Uses intelligent routing: web/news queries go to Tavily, docs/github/pdf queries go to Firecrawl, with fallback mechanisms to Exa when needed.

Compare MySearch-Proxy with