home / mcp / search-scrape mcp server

Search-Scrape MCP Server

Provides free web search and content scraping endpoints for AI assistants with local processing.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "devshero-search-scrape": {
      "command": "/path/to/mcp-server/target/release/mcp-stdio",
      "args": [],
      "env": {
        "SEARXNG_URL": "http://localhost:8888"
      }
    }
  }
}

Search-Scrape is a fast, local MCP server that provides free web search and scraping tools for AI assistants. It combines advanced search with optimized content extraction so your workflows can query the web and pull clean data without external keys or limits.

How to use

You will run the SearXNG-backed search engine locally, start the MCP server, and then connect your AI assistant to the MCP endpoint. Use the search tool to perform advanced web queries and the scrape tool to extract clean article content from URLs. Your assistant can then present results, summarize pages, or extract structured data from sources.

How to install

Prerequisites you need before installing this MCP server are Docker and Rust toolchain (cargo). Ensure Docker is running on your system.

1) Start the SearXNG search engine locally using Docker Compose.

# 1. Start SearXNG search engine
 docker-compose up searxng -d

2) Build the MCP server from source.

# 2. Build MCP server
 cd mcp-server && cargo build --release

3) Configure your AI assistant to connect to the MCP server using the following MCP configuration snippet.

{
  "mcpServers": {
    "search_scrape": {
      "command": "/path/to/mcp-server/target/release/mcp-stdio",
      "env": { "SEARXNG_URL": "http://localhost:8888" }
    }
  }
}

Available tools

search_web

Advanced web search with full SearXNG parameter support including engines, categories, language, safesearch, time_range, and pagination.

scrape_url

Optimized content extraction that removes ads and boilerplate, preserving headings and lists for clean article text.