Web Search is an MCP server that enables AI assistants to execute Google searches and view target web pages using bot detection avoidance techniques. It connects Model Context Protocol clients directly to web search and extraction workflows, bridging the gap between static language models and live online data. Developers, researchers, and automation engineers use this tool to provide autonomous agents with real-time web access without running into aggressive anti-scraping blocks or CAPTCHA challenges typical of standard automated browsing setups. Through this server, AI tools can look up current event details, extract text from documentation sites, cross-reference data across multiple search results, and inspect live page contents on demand. By incorporating evasive browsing strategies, the server ensures more reliable retrieval of search listings and web pages during continuous research sessions. Users can integrate this capability into their workflows to handle automated data gathering, verify facts, monitor external publications, and supply grounded context for queries that demand information beyond the training cutoff date.
Category: Browser & Web Automation
Tags: bot detection, google, scraping, web search
claude_desktop_config.json for Claude Desktop or the configuration settings inside Cursor. 4. Add a new entry for Web Search under the mcpServers configuration block specifying the runtime command, arguments, and required variables. 5. Restart your MCP client to complete registration and confirm that the search tools are available.Part of MCP Servers
Web Search is a Model Context Protocol server that allows AI applications to search Google and fetch web content while mitigating bot detection. It exposes search and browsing capabilities directly as tools for compatible AI agents, allowing models to query the live internet for relevant answers and read remote page content during conversations.
The server executes search queries through Google and fetches text and structure from target web pages. By implementing bot detection avoidance measures, it assists AI agents in reading web resources without getting blocked by standard anti-scraping filters, making it useful for automated web research and fact verification.
Web Search works with any client supporting the Model Context Protocol specification. Popular compatible clients include Claude Desktop, Cursor, and custom command-line or programmatic agent environments that support standard MCP server configurations over local transport mechanisms.
Detailed installation instructions should be referenced directly from the official GitHub repository at https://github.com/Claw256/mcp-web-search. Generally, you clone or download the server package, install any necessary environment dependencies, and register the execution command within your MCP client configuration file.