tavily-search-pro

What This Skill Does

The tavily-search-pro skill for OpenClaw integrates the powerful Tavily AI search engine directly into your agent workflow. Unlike traditional search engines, Tavily is purpose-built for AI agents, providing high-quality, relevant search results and synthesized answers that are ready for large language models to consume. This skill supports five distinct modes: Search (web/news/finance), Extract (URL content parsing), Crawl (deep website navigation), Map (sitemap discovery), and Research (multi-step comprehensive analysis). By using this skill, your OpenClaw agent can fetch real-time information, parse complex web documentation, or perform deep industry research with full citations, ensuring that your automated workflows are grounded in verified, up-to-date facts.

Installation

To integrate this skill, first ensure you have your Tavily API key ready. You can install the skill via the OpenClaw repository using the following command in your terminal:

clawhub install openclaw/skills/skills/shaharsha/tavily-search-pro

Once installed, add your API key to your configuration file:

{
  "env": {
    "TAVILY_API_KEY": "tvly-your-api-key-here"
  }
}

Use Cases

• Deep Research: Automate the gathering of information on complex topics, such as market trends, competitor analysis, or academic research, with built-in citation capabilities.
• Content Scraping: Extract clean, readable content from web articles or documentation pages, bypassing ads and boilerplate, and outputting the data in markdown format for direct use in reports.
• Financial Monitoring: Get focused, real-time insights into stock performance and market sentiments by utilizing the dedicated finance mode.
• News Summarization: Track industry shifts by querying news-specific databases, optionally using the --answer flag to get an LLM-synthesized summary of the latest events.

Example Prompts

1. "Perform a research query on the current status of carbon capture technology and provide a summary with citations."
2. "Extract the content from these three blog posts and summarize the key technical takeaways: [URL1], [URL2], [URL3]."
3. "What are the latest updates on the NVIDIA stock price and what are analysts predicting for the next quarter?"

Tips & Limitations

• Depth vs Cost: While 'advanced' depth provides better, more detailed results, it consumes two credits per request rather than one. Use it only when the query requires deep analysis.
• Domain Filtering: Use the --include-domains flag to improve signal-to-noise ratios when researching within specific, trusted sources like academic databases or official documentation sites.
• Rate Limiting: Be aware of your Tavily account limits. Large-scale crawling and high-frequency research tasks may consume your monthly credit allowance quickly. Always check the API documentation for usage tiers.