Performs web searches through Tavily's API with results optimized for LLM consumption, returning content snippets, relevance scores, and metadata in structured JSON. Handles domain filtering, time ranges, and different search depths from ultra-fast to advanced precision modes. Most useful when you need current information or want to gather sources before deeper research. The LLM-optimized results format means less parsing work compared to raw web scraping, and the domain filtering helps focus on trusted sources. Install requires the Tavily CLI and API authentication, but the structured output makes it worth the setup for regular web research tasks.
npx skills add https://github.com/tavily-ai/skills --skill tavily-search