Tavily Search - Web Search for OpenClaw¶

Publisher: framix-team | Installs: 797 | Category: Search

Integrates the Tavily Search API into OpenClaw - purpose-built search for AI agents. Returns clean, structured results with source attribution, topic classification, and content extraction. Far better than raw Google scraping for agent consumption.

Prerequisites¶

OpenClaw installed and running
Tavily API key (free tier: 1,000 searches/month)
Node.js 18+
npx available

Installation¶

npx skills add framix-team/openclaw-tavily/tavily-search

Configuration¶

Get your API key from tavily.com and configure:

{
  "tavily": {
    "apiKey": "${TAVILY_API_KEY}",
    "searchDepth": "advanced",
    "maxResults": 10,
    "includeAnswer": true,
    "includeRawContent": false,
    "includeImages": false,
    "searchDomain": "general"
  }
}

export TAVILY_API_KEY="tvly-..."

Setting	Default	Description
`searchDepth`	`advanced`	`basic` (fast) or `advanced` (comprehensive)
`maxResults`	`10`	Max results per search (1-20)
`includeAnswer`	`true`	Include AI-generated answer summary
`includeRawContent`	`false`	Include full page text (token-heavy)
`includeImages`	`false`	Include image results
`searchDomain`	`general`	`general` or `news` for time-sensitive queries

Capabilities¶

AI-Optimized Search¶

Searches across the web with a single API call
Returns cleaned, structured results (no HTML noise)
AI-generated answer synthesis from top results
Source attribution with URLs and relevance scores

Topic Extraction¶

Classifies search results by topic
Identifies key entities and relationships
Groups related results for consumption

Content Extraction¶

Extracts clean text from result pages
Removes ads, navigation, and boilerplate
Returns structured content for agent reasoning

CLI Reference¶

Command                       Description
──────────────────────────────────────────────
tavily-search <query>         Run a web search
tavily-search news <query>    Search recent news
tavily-search extract <url>   Extract content from URL
tavily-search topics <query>  Get topic breakdown
tavily-search answer <query>  Get AI-generated answer only

Example Usage¶

# Basic search
tavily-search "latest Hermes agent updates 2026"

# News search (time-sensitive)
tavily-search news "OpenAI GPT-5 announcement"

# Extract content from a specific page
tavily-search extract "https://hermes-agent.nousresearch.com/docs"

# Get topic analysis
tavily-search topics "autonomous AI agent deployment patterns"

CorpusIQ Use Cases¶

Market research: Agent searches for competitor updates, industry trends, pricing changes
Content curation: Gather sources for blog posts, social media content, newsletters
Lead research: Research prospect companies before outreach
Technical research: Search for solutions to agent-deployment problems
News monitoring: Track mentions of CorpusIQ and ecosystem partners

Troubleshooting¶

API key not recognized¶

# Verify key is set
echo $TAVILY_API_KEY

# Test directly
curl -s "https://api.tavily.com/search" \
  -H "Content-Type: application/json" \
  -d '{"api_key":"'$TAVILY_API_KEY'","query":"test"}'

Search returns empty¶

Check searchDomain - news only returns recent content
Try searchDepth: basic - advanced may be slower but more thorough
Verify API key has quota remaining

Rate limiting¶

# Check remaining quota
curl -s "https://api.tavily.com/usage" \
  -H "Content-Type: application/json" \
  -d '{"api_key":"'$TAVILY_API_KEY'"}'

Free tier: 1,000 searches/month. Upgrade for higher limits.

← Tavily Search Setup Guide | Discovery Page →

↑ Skills Catalog

Part of the Hermes Skills Library - curated by CorpusIQ. Content remains attributed to original authors and repositories. CorpusIQ - one MCP endpoint, all your business tools.

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects