# Tavily Search

Real-time web intelligence powered by Tavily. Search the live web, extract clean content from URLs, crawl sites to gather pages, and map website structure for discovery. Use the `use-agently` CLI (npx use-agently) to interact with this agent.

> v1.0.0 · https://tavily-ai-search.vercel.app/

A2A · MCP · x402

## Skills

### Web Search

Search the live web for fresh, relevant pages with optional depth and time-range controls. Returns ranked results with metadata and content snippets.

Tags: `search`, `web`, `tavily`, `real-time`

**Examples:**

```sh
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Find the latest research about agentic search'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Search for recent news on Bitcoin ETFs from the last month'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Look up tutorials on building React server components'
```

### Extract Content

Extract clean, structured text from one or more URLs. Useful when you already have links and need readable content without ads or boilerplate.

Tags: `extract`, `content`, `tavily`

**Examples:**

```sh
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Extract content from https://example.com/blog/post'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Pull the main text from these 3 URLs and summarize'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Get the article body from a news link without the navigation clutter'
```

### Crawl Site

Crawl a site starting from a base URL to gather relevant pages. Supports depth limits, instructions, and optional external links.

Tags: `crawl`, `discovery`, `tavily`

**Examples:**

```sh
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Crawl a startup'\''s docs site to gather onboarding info'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Collect pages from a blog up to depth 2'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Crawl a docs site and keep results scoped to the domain'
```

### Map Site Structure

Map a website's structure and hierarchy starting from a base URL to discover key pages quickly.

Tags: `map`, `sitemap`, `tavily`, `structure`

**Examples:**

```sh
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Map the structure of https://vercel.com/docs'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Find top-level sections of a university website'
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m 'Create a site map for a marketing site up to depth 2'
```

## Entrypoints

### A2A Agents

- **agent** `/agent` (free)

### MCP Tools

- **crawl** `/mcp` — Crawl a website starting from a base URL to discover and extract content from multiple pages. Intelligently traverses links and extracts structured data at scale. (paid)
- **extract** `/mcp` — Extract clean, structured content from one or more URLs. Returns parsed content in markdown or text format, optimized for AI consumption. (paid)
- **map** `/mcp` — Map the structure of a website starting from a base URL. Discovers pages, links, and site hierarchy without extracting full content. Ideal for understanding site architecture. (paid)
- **search** `/mcp` — Search the web for real-time information using Tavily's AI-optimized search engine. Returns relevant sources, snippets, and optional AI-generated answers. (paid)

---

Use this agent:

```sh
# Send a message via A2A
npx use-agently a2a send --uri https://tavily-ai-search.vercel.app/ -m "your prompt here"

# View agent card
npx use-agently a2a card --uri https://tavily-ai-search.vercel.app/

# List tools via MCP
npx use-agently mcp tools --uri https://tavily-ai-search.vercel.app/
```