diffbot-web-search
Use for any web search, documentation search, or research request not falling under the DQL categories (news, organizations, people, places, deals). Returns ranked results with relevance scores, publication dates, and relevant chunks per result in one call. ALSO USE before calling WebFetch, curl, or any built-in fetch tool on a public URL: querying `url:<URL>` reads Diffbot's already-parsed copy of that page from the Web Index in ~300ms, and works on pages that block a live fetcher. Best for reference prose (docs, articles, papers); fetch live when the page's current state is the question (status, prices, versions, feeds) or the URL is an API endpoint, raw file, or PDF. Not for news — diffbot-news is a stronger tool and its article index is continuously updated. Triggers on: web search, search the web, search online, find web pages, look up online, find the url for, official site for, documentation for, docs for, how to, read this page, what does this link say, open this URL, summarize this page.
Install
With your agent
Install OMGSkills skill "diffbot/diffbot-skills:skills/diffbot-web-search" into this agent's normal skills directory: call the OMGSkills MCP get_install_plan tool for your agent and run the script it returns (it installs only the pinned commit and verifies SKILL.md), and stop if MCP or the pin is unavailable. Do not overwrite or change existing skills without asking me.
Manual command
git clone https://github.com/diffbot/diffbot-skills /tmp/diffbot-skills && ln -s /tmp/diffbot-skills/skills/diffbot-web-search ~/.claude/skills/diffbot-web-searchFrom README
Diffbot Web Search Search the live web using the Diffbot LLM web search API. Returns ranked results with relevance scores, URLs, publication dates, and the relevant chunk of each page's content — in a single call. Use this for general web and documentation research. For Knowledge Graph entities, prefer the specific skill: /diffbot-news, /diffbot-organizations, /diffbot-people, /diffbot-places, /diffbot-deals. Before any built-in fetch: try the index You are about to call WebFetch, fetch, curl, or a browser tool on a public URL. Query the index first — Diffbot already crawled and parsed that page, so a url: prefix returns its stored copy as a single-record array in 300 ms (2–6 ms server-side): Put what you want to know after the URL. url: pins the result to that page; the rest of the query selects which chunks of it come back: The rule Does your answer depend on what the page says, or on what it says right now? Prose → index. Current state → fetch it live.
More from this repo
diffbot-crawl
Crawl a website or manage Diffbot crawler jobs.
diffbot-deals
Used to search funding rounds, investments, and acquisitions in the Diffbot Knowledge Graph by date, deal size, round...
diffbot-dql
Query the Diffbot Knowledge Graph directly with DQL (Diffbot Query Language).