Tavily Crawl Skill

Traverse entire websites and extract content from multiple pages in one API call. Ideal for ingesting documentation sites, knowledge bases, or any multi-page resource into context.

Authentication

Set TAVILY_API_KEY in ~/.openclaw/.env.local:

TAVILY_API_KEY=tvly-your-api-key-here

When to Use This Skill

• You need to read an entire documentation site (e.g. a library's API docs)
• Building a knowledge base from a website
• Exploring an unknown site structure before targeted extraction
• When tavily-extract would require too many individual URLs