> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tembo.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Tavily

> Search, extract, map, and crawl public web content through Tavily MCP.

Search, extract, map, and crawl public web content through Tavily MCP.

## Prerequisites

You need access to your Tembo organization and access to Tavily search tools. Review the provider permissions before granting access to your data.

## Installation

1. Open **Settings → Integrations** in Tembo.
2. Select **Tavily** and start the installation.
3. Complete the provider authorization if prompted. Review the requested access and select the intended account or workspace.
4. Confirm the connection is ready before starting a new session.

The catalog server endpoint is `https://mcp.tavily.com/mcp`.

If authorization requires a pre-registered OAuth client, follow the [MCP authentication instructions](/integrations/mcp#advanced) and register the callback shown in the connection dialog.

## Usage and verification

Start a new session and ask Tembo to search for a non-sensitive public topic. Confirm the tools are available and that any returned account data belongs to the account you intended to connect.

## Access and troubleshooting

Do not put internal secrets in search queries. Review the scope before asking the agent to crawl sites.

If the connection remains pending, return to **Settings → Integrations** and finish authorization. If tools do not appear, confirm the connection was saved and start a new session. See [MCP troubleshooting](/integrations/mcp#advanced) for authentication and connectivity checks.