A specialized search engine skill designed to provide Large Language Models with clean, structured web data and synthesized AI answers.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install tavily
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install tavily using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Tavily AI Search?
Tavily AI Search is a sophisticated search tool built specifically for AI agents and LLM applications. Unlike traditional search engines that return messy HTML and ads, this skill delivers structured, noise-free content that is immediately ready for processing by an AI model. By integrating this into your collection of Openclaw Skills, you enable your agent to conduct high-speed factual lookups or deep-dive technical research with ease.
The skill supports multiple operational modes, including a basic mode for rapid responses and an advanced mode for comprehensive analysis. It excels at extracting clean content and generating automated answer summaries, making it an essential component for any developer building autonomous research workflows or data-gathering agents.
Tavily AI Search Use Cases
- Performing comprehensive research on complex technical topics like quantum computing or software architecture.
- Tracking current events and breaking news using the time-sensitive news topic filter.
- Restricting searches to authoritative documentation or academic domains like GitHub, arXiv, or official project sites.
- Extracting raw HTML content for deep-dive sentiment analysis or custom data scraping projects.
- Generating quick, synthesized summaries of search results to provide immediate answers without manual browsing.
How Tavily AI Search Works
- The AI agent initiates a query through the Openclaw Skills interface, specifying parameters like search depth and topic focus.
- The skill applies domain filters to include or exclude specific websites, ensuring a high signal-to-noise ratio.
- The Tavily API executes the search, crawling the web or specific news sources based on the user's requirements.
- Search results are parsed to extract clean text snippets, relevance scores, and optional raw HTML or images.
- If requested, an AI model synthesizes the findings into a concise answer summary.
- The final structured payload is returned to the agent, providing it with grounded, real-world information.
Tavily AI Search Setup
To integrate this search capability into your Openclaw Skills workflow, follow these installation steps:
- Install the Tavily Python SDK:
pip install tavily-python
- Obtain an API key from the Tavily dashboard and set it as an environment variable:
export TAVILY_API_KEY="tvly-YOUR_API_KEY_HERE"
- Alternatively, add the configuration to your agent settings:
{
"skills": {
"entries": {
"tavily": {
"enabled": true,
"apiKey": "tvly-YOUR_API_KEY_HERE"
}
}
}
}
Tavily AI Search Data Schema & Taxonomy
The skill organizes its output into a highly structured format optimized for context windows. This structure is a hallmark of quality within the Openclaw Skills ecosystem.
| Attribute | Description |
|---|---|
answer |
A synthesized summary of the search results for quick comprehension. |
results |
An array containing title, URL, and optimized content snippets. |
score |
A numerical relevance score (0-1) indicating the quality of each result. |
raw_content |
The full HTML content of the page (if requested for deep analysis). |
images |
A list of URLs for relevant visual content found during the search. |
Tavily AI Search Advanced Features
- Advanced Search Depth: A multi-pass search process for complex queries requiring maximum coverage.
- News-Specific Filtering: A dedicated mode that prioritizes content published within the last 7 days.
- Domain Inclusion/Exclusion: Strategic control over source quality by whitelisting trusted sites or blacklisting content farms.
- Programmatic Chaining: Ability to pipe search results into other CLI tools or automation scripts via JSON output.
- Raw HTML Extraction: Capability to bypass snippets and retrieve full page source for specialized parsing.
SKILL.md
Loading
Related Openclaw Skills
Trading Coach
An AI trading coach that converts broker CSV exports into professional review reports with 8-dimensional scoring and deep pattern insights.
RSS AI Reader
An intelligent RSS and Atom feed reader that uses LLMs to generate summaries and delivers them through multiple notification channels.
Options Analyzer
A professional-grade options strategy analyzer providing real-time market data, Greeks calculation, and intelligent strategy recommendations.
Clawdio
Clawdio provides auditory perception for AI agents, converting human audio into structured data, semantic reports, and machine-readable markdown.
Instagram Reel Downloader for WhatsApp
Download Instagram Reels via browser automation and send them as WhatsApp-ready videos using Openclaw Skills.
Recipe Video Extractor
An AI-powered tool that converts video URLs into structured recipes by analyzing captions, comments, and web data.
Featured*
