Search and Extract Web Content
Adds web search, URL content extraction, and site crawling to Claude Code conversations via the Tavily API.
Why it matters
Leverage the Tavily API to perform web searches, extract specific content from URLs, and crawl websites for up-to-date information.
Outcomes
What it gets done
Conduct web searches for current information.
Extract content from specified URLs.
Crawl websites to gather data.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/ag-tavily-web | bash Overview
Tavily Web
A community skill that connects Claude Code to the Tavily API for web search, URL content extraction, and website crawling, installed via npx and configured with a Tavily API key. Use when a Claude Code conversation needs current information, page content from a specific URL, or crawled site content beyond the model's static training data.
What it does
This skill wires Claude Code into the Tavily API to add three capabilities in one package: web search for current information, content extraction from specific URLs, and crawling of websites. It installs as a community skill via npx skills add -g BenedictKing/tavily-web and, once a Tavily API key is configured, is invoked naturally inside Claude Code conversations rather than through a dedicated command syntax.
Inputs and outputs
Input is a search query, a URL to extract, or a site to crawl, plus a configured Tavily API key (the skill's own best-practice guidance recommends setting this via environment variables rather than hardcoding it). Output is search results, extracted page content, or crawled site content, surfaced directly back into the Claude Code conversation.
npx skills add -g BenedictKing/tavily-web
Integrations
Built entirely on the Tavily API for its search, extraction, and crawl operations. The skill's own documentation lists it alongside related research-oriented skills - context7-auto-research, exa-search, firecrawl-scraper, and codex-review - indicating it's designed to slot into a broader research and verification workflow rather than stand alone as the only web-access tool in a session.
Who it's for
Developers and researchers using Claude Code who need live web search, page-content extraction, or basic site crawling inside their coding or research sessions, rather than relying solely on the model's static training data for anything time-sensitive or URL-specific.
FAQ
Common questions
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.