Skip to content
Get packs

Tavily Web Search Best Practices

Teaches the agent to build Tavily web search, extraction, crawling and research integrations correctly.

Source Download skill
Repository
andrewyng/context-hub
Stars
13,984
Licence
MIT
Files
7 files

What it does

Tavily is a web search service built for AI apps. This skill gives the agent a reference guide for using Tavily's Python and JavaScript libraries: how to set up a client, which method to use (search, extract, crawl, map or research), and the key options for each. When you ask for a web-search or RAG feature, the agent follows these patterns instead of guessing at the API, and it can read deeper reference files for framework integrations such as LangChain and LlamaIndex.

Great for

  • Developers adding live web search or page extraction to an AI agent or RAG pipeline
  • Teams choosing between Tavily's search, extract, crawl and research methods
  • Engineers wiring Tavily into frameworks like LangChain, LlamaIndex, CrewAI or the Vercel AI SDK

Skip it if

  • Non-developers who only want to search the web themselves; this is a coding reference, not a search tool
  • Projects that must keep data out of third-party services, since every query and crawled page is sent to Tavily's API
  • Anyone without a Tavily account and API key, because the code examples will not run

Try asking

  • “Add Tavily web search with domain filtering to my LangChain agent”
  • “Write a Python function that extracts clean text from a list of documentation URLs using Tavily”
  • “Set up a Tavily research call with structured JSON output and poll for the result”
  • “Crawl a docs site for API pages and feed the results into my RAG index”
Step-by-step guide: how to use Tavily Web Search Best PracticesInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 7 original files, the MIT licence and an attribution note.

    tavily-best-practices.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip tavily-best-practices.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip tavily-best-practices.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Provide your TAVILY_API_KEY as an environment variable rather than pasting it into prompts or code.
  • Name the framework and language you use so the agent picks the matching integration and install commands.
  • Review generated code before running it, especially crawl limits, since crawling can send many requests and consume API credits.

What's inside

  • references/crawl.md
  • references/extract.md
  • references/integrations.md
  • references/research.md
  • references/sdk.md
  • references/search.md
  • SKILL.md

Redistributed unmodified from andrewyng/context-hub under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.