Overview
An MCP server implementation that integrates the SearXNG API, providing web search and URL content reading capabilities. It exposes two main tools: searxng_web_search for paginated web searches and web_url_read to fetch a URL and extract content into Markdown with advanced options such as section filtering and headings. It supports intelligent caching with TTL to reduce redundant requests, and offers search result controls like pagination, time range filtering, language selection, and safesearch. Configuration is via environment variables (SEARXNG_URL required, optional AUTH_USERNAME/AUTH_PASSWORD, USER_AGENT, HTTP_PROXY/HTTPS_PROXY, NO_PROXY). Deployment options include NPX, NPM, and Docker, with an HTTP transport mode (MCP_HTTP_PORT) for HTTP endpoints. The project emphasizes TypeScript, unit tests, and MCP inspector compatibility, and includes testing/eval commands.
Features
Web Search
Execute web searches with pagination, returning results from external services (queries, news, articles).
URL Content Reading
Read and convert content from a URL to Markdown with advanced options like section extraction, heading extraction, and paragraph ranges.
Intelligent Caching
Cache URL content using TTL to improve performance and reduce redundant requests.
Pagination
Control which page of results to retrieve for search results.
Time Filtering
Filter results by time range (day, month, year).
Language Selection
Filter results by preferred language.
Safe Search
Control content filtering level for search results.
