Tavily Web Search

# Tavily Web Search Search the web using the Tavily API with AI-powered results and content extraction from web pages. AI-Powered Web Search — Queries the Tavily API to return intelligent, context-aware search results from the web. Content Extraction — Retrieves both snippets and full page content from matched results. Domain Filtering — Restrict results to specific domains or exclude unwanted sources. Topic-Based Search — Target general web results or focus specifically on news. Configurable Depth — Choose between basic (fast) and advanced (deeper) search modes. ## How It Works 1 Step 1 API Key Configuration Set your TAVILY_API_KEY environment variable to authenticate with the Tavily service. 2 Step 2 Query Submission Pass a search query to the skill via the CLI or ClawdBot tool invocation. 3 Step 3 Tavily API Call The skill sends your query to the Tavily API with your chosen options (depth, domains, topic). 4 Step 4 Result Processing Tavily returns AI-ranked results with snippets and optionally full page content. 5 Step 5 Output Delivery Results are returned to ClawdBot for use in downstream reasoning or display. ## Key Features - **AI-Powered Results**: Tavily ranks and filters results using AI for higher relevance. - **Content Extraction**: Returns full content from matched web pages, not just snippets. - **Domain Filtering**: Include or exclude specific domains with --include-domains / --exclude-domains . - **Topic Mode**: Use --topic news for news-focused searches or --topic general for broad web queries. - **Search Depth Control**: Switch between basic and advanced depth for speed vs. thoroughness trade-offs. - **Result Count**: Configure the number of results returned with --max-results . ## Requirements - **TAVILY_API_KEY**: Required. Powers all web search and content extraction. Get a free API key at tavily.com . ## Use Cases - **Real-Time Research**: Let ClawdBot look up current events, news, or technical documentation on demand. - **Fact Checking**: Verify claims or retrieve up-to-date information beyond the model's training cutoff. - **News Monitoring**: Use topic-based news search to stay informed on specific subjects. - **Domain-Specific Lookup**: Narrow searches to trusted sources for more reliable results. - **Content Summarization**: Extract full page content and pass it to ClawdBot for summarization or analysis. ## Installation Install via: `npx clawhub@latest install tavily-search`

Instalação

Habilidades relacionadas