tavily

What This Skill Does

The Tavily AI search skill provides a specialized interface for AI agents to query the internet. Unlike standard search engines that prioritize SEO-optimized advertisements and cluttered layouts, Tavily is purpose-built to extract clean, relevant, and context-rich snippets directly from the web. This tool allows your OpenClaw agent to perform real-time research, verify facts, or stay updated on current events without the noise typically associated with raw web scraping. It supports multiple search modes, including a deep research mode for complex investigations and a dedicated news aggregator for trending topics.

Installation

To integrate this capability into your environment, use the OpenClaw hub command. Ensure you have your TAVILY_API_KEY ready, which can be obtained via the official Tavily developer portal. Run the following command in your terminal:

clawhub install openclaw/skills/skills/cp33333333333/tavily-search1

Once installed, set your API key as an environment variable to ensure the agent can authenticate and authorize requests correctly.

Use Cases

This skill is highly versatile for any agent-based workflow. Common use cases include: 1. Research & Synthesis: Conducting comprehensive literature reviews or gathering background information on new technologies. 2. Real-time News Monitoring: Tracking live events as they unfold by using the specialized news topic flag. 3. Fact Checking: verifying claims against authoritative web sources to minimize hallucinations in AI-generated text. 4. URL Analysis: Leveraging the extract tool to parse specific content from web articles without being hindered by boilerplate web content like navigation bars or footers.

Example Prompts

1. "Perform a deep search on the latest advancements in solid-state battery technology and summarize the key industry challenges."
2. "Search for the most recent news regarding global market shifts in the last 3 days and explain how they might affect the tech sector."
3. "Extract the content from https://example.com/research-paper and provide a list of the 5 most important conclusions drawn by the authors."

Tips & Limitations

The --deep search mode is significantly more resource-intensive and slower; reserve this for complex queries requiring exhaustive research. For general fact-checking, stick to the default settings to maintain agent responsiveness. Always remember that while Tavily cleans content, the quality of the output depends on the source's original clarity. If you encounter errors, verify that your TAVILY_API_KEY has sufficient quota and that your internet connection permits access to the Tavily API endpoints.