5 Downloads
5 Downloads
jorbraken/web-searchGive LLMs tools to search the web, fetch pages, and extract clean, cited Markdown content.
search_web(query, count?) — searches the web and returns title, url, snippet, source,
and published_date for each result. Uses a local SearXNG instance by default, falling back to
the Brave Search API if SearXNG is unreachable or errors out.fetch_page(url, offset?) — fetches a page, strips navigation/ads/scripts, and returns the main
content as Markdown along with citation metadata (title, url, source, published_date).
Long pages are returned in chunks; pass offset to continue reading a page from where the
previous chunk left off.text/html pages; PDFs, images, and other content types are not supported.@mozilla/readability and may vary on non-article pages (e.g.
heavily JS-rendered SPAs).jorbraken/web-searchGive LLMs tools to search the web, fetch pages, and extract clean, cited Markdown content.
search_web(query, count?) — searches the web and returns title, url, snippet, source,
and published_date for each result. Uses a local SearXNG instance by default, falling back to
the Brave Search API if SearXNG is unreachable or errors out.fetch_page(url, offset?) — fetches a page, strips navigation/ads/scripts, and returns the main
content as Markdown along with citation metadata (title, url, source, published_date).
Long pages are returned in chunks; pass offset to continue reading a page from where the
previous chunk left off.text/html pages; PDFs, images, and other content types are not supported.@mozilla/readability and may vary on non-article pages (e.g.
heavily JS-rendered SPAs).http://192.168.66.2:8888).count isn't specified.fetch_page call.http://192.168.66.2:8888).count isn't specified.fetch_page call.