tavily-search

The tavily-search skill is a high-performance tool designed for AI agents to perform real-time web research and gather accurate, current information. By leveraging the Tavily API, the agent can bypass the constraints of static training data and access the latest online developments, news, and technical documentation. This skill is ideal for researchers, analysts, developers, and knowledge workers who require up-to-date data to inform decision-making processes. It excels in scenarios where temporal relevance is critical, such as monitoring industry trends, researching specific technical topics, or fetching live news updates. Unlike general-purpose web scrapers, this tool is optimized for research-oriented queries, providing summarized content, multi-source synthesis, and intelligent filtering of web data.

• Performs real-time web searches across diverse domains and news outlets

• Extracts structured content and full-page information for deeper analysis

• Supports configurable search depth including basic and advanced research modes

• Offers customizable parameters for result counts, AI-generated synthesis, and raw content retrieval

• Integrates seamlessly with local environment configurations or standard API key management workflows

• Requires a valid Tavily API key and network access to the Tavily API endpoint

• Best suited for search-intensive tasks; for simple URL fetching, consider using a dedicated web_fetch tool instead

• API usage is subject to the rate limits and billing tiers associated with your Tavily account

• Input consists of natural language search queries and optional parameters like search_depth or max_results; output typically returns search result objects or AI-summarized answers

• While designed for research, always verify critical information against primary sources when dealing with sensitive or highly technical data

• Ensure the API key is securely managed via environment variables or the openclaw.json configuration file to prevent unauthorized access