Performs web searches and extracts page content using the Tavily API.
npx skills add https://github.com/srgaba/open-claw --skill tavily-srgaba
Please help me install this Agent Skill. Skill: tavily Source: https://github.com/srgaba/open-claw/tree/main/project/extensions/tavily/skills/tavily Command: npx skills add https://github.com/srgaba/open-claw --skill tavily-srgaba
web_search or fine-grained searches with tavily_search using depth, topic, domain, and time-range filters. - Content Extraction: Use tavily_extract to pull clean markdown from up to 20 URLs, including JavaScript-rendered pages, with query-focused chunking. - Use Case: Research a breaking news topic by running tavily_search with the news topic and a week time range, then extract full article text from the top URLs with tavily_extract. ## Quick Start Use the tavily skill to search for the latest news on renewable energy policy and extract the content of the top three result URLs.High-intent search queries and answers about installing and using this skill.
Use web_search for straightforward queries with just a query string and result count. For more control, use tavily_search to set search depth, topic, domain filters, time ranges, and AI-generated answer summaries.
Use tavily_extract with an array of up to 20 URLs. It returns clean markdown and handles JavaScript-rendered pages. Add a query parameter with chunks_per_source to get only the most relevant sections instead of full pages.
Basic depth is faster and suited for general-purpose queries, while advanced depth is slower but delivers the highest relevance. Use advanced for precision tasks, specific facts, or detailed research.
Yes, tavily_extract handles JavaScript-rendered pages. Start with basic extract depth for simple pages, and switch to advanced depth for JS-heavy SPAs, dynamic content, tables, or embedded content.
Search queries should stay under 400 characters and return 1-20 results. Extraction accepts a maximum of 20 URLs per request, so larger URL lists must be batched into multiple calls.
adrianliechtiAggregate multi-source research context into a structured brief with attribution and confidence scoring.
T-0-coSearch Notion workspaces and synthesize findings into structured reports with citations.
tocadaoncamodConduct structured online research with cross-verified, cited sources.
faisalanjumRetrieve raw web search results with URLs and snippets from Perplexity AI.