BenedictKing built this to give Claude Code direct access to Tavily's web search API for pulling current information, extracting content from URLs, and crawling sites. You'll want it when you need Claude to look up something that's not in its training data or grab content from specific pages during development. Setup is straightforward with an API key in your environment variables, then you just ask Claude naturally. The repo lists context7-auto-research, exa-search, and firecrawl-scraper as related options, so if you're comparing web search integrations, those are worth checking too. Solid utility skill if you're already using Tavily or want to add real-time web data to your Claude workflows.
npx skills add https://github.com/davila7/claude-code-templates --skill tavily-web