Wraps the Serper API to bring Google search results and web scraping directly into Claude Desktop. Exposes two core operations: google_search for retrieving organic results, knowledge graphs, and related searches with support for advanced operators like site: and filetype:, and scrape for extracting clean text and markdown from any webpage. You'd reach for this when you need Claude to pull current web information or research topics beyond its training data. Also integrates with Cline and Cursor if you're working in those environments. Requires a Serper API key but handles all the complexity of search result parsing and content extraction.
claude mcp add --transport stdio marcopesani-mcp-server-serper uvx mcp-server-serper