Tavily

tavily-best-practices

Build production-ready Tavily integrations with best practices baked in. Reference documentation for developers using coding assistants (Claude Code, Cursor, etc.) to implement web search, content extraction, crawling, and research in agentic workflows, RAG systems, or autonomous

tavily-cli

Install and authenticate the Tavily CLI (`tvly`) in a coding-agent terminal. Use only when the user asks to install tvly, run tvly login, or check CLI status, or when a terminal session has no callable Tavily MCP tools. Not for ordinary web search, extraction, mapping, crawling,

tavily-crawl

Collect content from many pages on the same site. Use when the user wants to crawl, download docs, extract a whole section such as /docs, bulk-extract pages, or says "get all the pages" or "extract everything under".

tavily-extract

Extract clean content from known URLs. Use when the user supplies one or more URLs and wants the page text, says "extract", "grab the content from", "pull the text from", "get the page at", or "read this webpage".

tavily-map

Discover URLs on a known website without extracting page content. Use when the user wants the site structure, to find a specific page on a large site, or says "map the site", "list the pages", "what URLs are on", or "find the docs URL".

tavily-research

Produce cited multi-source research. Use when the user wants a deep report, comparison, market analysis, or literature review, or says "research", "investigate", "analyze in depth", "compare X vs Y", or "what does the market look like for". For a quick fact, use tavily-search ins

tavily-search

Search the web for current sources and snippets. Use when the user wants to search, look up, find articles, get recent news, or discover sources, and does not already have a URL. Triggers: "search for", "look up", "find me", "what's the latest on", "find articles about".