tavily

Retrieve AI-optimized web search results with configurable depth and topic.

Updated Apr 5, 2026
One-click install
npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill tavily-dingdyan
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tavily
Source: https://github.com/dingdyan/openclaw-workspace-v2/tree/main/skills/tavily-search
Command: npx skills add https://github.com/dingdyan/openclaw-workspace-v2 --skill tavily-dingdyan

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill helps you quickly find concise, relevant information on the web so you can answer questions without manually browsing and filtering results.

Core Features & Use Cases

  • AI-optimized web search via Tavily API: Retrieves sources and an optional AI-generated answer tailored for agent workflows.
  • Configurable result depth and topic focus: Supports basic vs deep search and topic scoping (e.g., general vs news).
  • Optional news time window: Limits news searches to recent items using a days parameter.

Quick Start

Run the Tavily search for “latest developments in EU AI regulation” and return up to 10 high-relevance sources.

Frequently Asked Questions about tavily

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get AI-optimized answers from web search for my agent workflow?

AI-optimized web search retrieves concise, relevant sources and optional generated answers via POST requests to the Tavily API, tailored for agent research workflows. You configure parameters like search depth and result count to get fast, relevant information.

Do I need an API key to retrieve web search results for AI agents?

Yes, you need a valid TAVILY_API_KEY environment variable to retrieve web search results. This key authenticates POST requests to the search and extract endpoints, returning sources and optional AI-generated answers for your queries.

How can I limit news research results to only recent articles?

To limit news research to recent articles, use the topic parameter set to news and apply the days parameter. This constrains the web search results to a specific recent time window, returning only timely news sources.

What's the best way to control the depth of information retrieval for research?

Control information retrieval depth by adjusting the search_depth parameter between basic and deep options. Deep search returns more comprehensive sources, while basic search provides faster, high-relevance web results for general research queries.

Can I extract specific answer content alongside general web search sources?

Yes, you can extract specific answer content alongside general web search sources. The process uses optional extract endpoints to retrieve AI-generated answers tailored to your query, configurable with parameters like max_results and topic scoping.