Metabase

Search and data

Metabase for AI agents. Managed auth, ready-made tools.

Browse dashboards and questions, run saved and new queries, and save questions in Metabase on behalf of each of your users.

About Metabase

With the Metabase connector, your agent works in each user’s own Metabase. It can list dashboards and saved questions, run a question and get its results, query a connected database and save a new question, using only the tools your project allows.

Metabase is an open source business intelligence tool for dashboards and data questions.

Tools

What your agent can do in Metabase.

Each tool is a capability the model can call. You decide which ones a project allows, and every call runs with the end user's own Metabase connection.

  • List dashboards

    Read
    metabase_list_dashboards

    List the dashboards the user can see.

  • Get a dashboard

    Read
    metabase_get_dashboard

    Get a dashboard and the questions on it.

  • List questions

    Read
    metabase_list_questions

    List saved questions.

  • Run a question

    Read
    metabase_run_question

    Run a saved question and get its results.

  • List databases

    Read
    metabase_list_databases

    List the databases connected to Metabase.

  • Run a query

    Read
    metabase_run_query

    Run a query against a connected database and get the results.

  • Save a question

    Write
    metabase_create_question

    Save a new question with its query.

Control

Limits you can set.

You decide what your agent may do in Metabase, and Agent200 runs every call within those limits.

  • Turn single tools on or off

    Let the agent read dashboards and run questions, and switch off metabase_create_question so it can never save a question.

  • Restrict what the agent reaches

    Choose the endpoints and scopes a project may use, for example saved questions but not new database queries. A call outside them is refused.

Get started

Connect Metabase
in three steps.

  1. 1

    Set up the SDK

    Add the Agent200 SDK to the codebase you already have and authenticate with your API key. The key covers every connector enabled for the project.

    agent.js
    Conceptual example
    Set up the SDK with the Agent200 API key
    // One Agent200 API key covers every connector enabled for the project.
    const agent200 = new Agent200({
      apiKey: process.env.AGENT200_API_KEY,
    });
  2. 2

    Let your user connect Metabase

    Each end user authorizes their own Metabase account once. Agent200 runs the OAuth flow, handles the callback, stores the tokens and refreshes them.

    connect.js
    Conceptual example
    Let an end user connect their account (managed OAuth)
    // Each end user authorizes their own Metabase account, once.
    const connection = await agent200
      .user("user_123")
      .connect("metabase");
    
    if (connection.authorizationUrl) {
      // Show this link to your user; Agent200 handles the rest.
      console.log(connection.authorizationUrl);
    }
  3. 3

    Run with the Metabase tools

    Name the user, the model and the tools. The model decides which Metabase tools to call, and Agent200 executes them with that user's connection.

    run.js
    Conceptual example
    Run a request with the connector's tools
    const result = await agent200
      .user("user_123")
      .run({
        model: "openai/gpt-5.6-sol",
        input: "What does the weekly revenue dashboard show",
        tools: ["metabase"],
      });
    
    console.log(result.output);

Access

Sign-in and permissions.

  • Who signs in

    Each of your users, once, by connecting their own Metabase account. Each connection is their own.

  • What access is requested

    The connection acts with that user's access in Metabase, such as reading dashboards and running questions. Users can revoke it later from their Metabase account settings.

For agents

Instructions for AI agents.

Using a coding agent? Paste this brief into it, or give it the address of this page. It describes the Metabase connector and how to use it.

Agent brief
Connector: Metabase, through Agent200. In code: "metabase".
What it does: Browse dashboards and questions, run saved and new queries, and save questions in Metabase on behalf of each of your users.
Page: https://www.agent200.com/connectors/metabase/

1. Get an Agent200 API key. If the developer has none, send them to https://agent200.com/platform/apikeys/ and ask them to choose "Create an API key". Read the key from the AGENT200_API_KEY environment variable. Never ask for it in chat, and never write it into code, prompts or logs.
2. Sign-in is managed OAuth. Each end user approves access once, on Metabase's own page, through the authorization link Agent200 returns. Never ask a user for a Metabase password or token in chat.
3. Give every request a stable id for the end user. Agent200 uses only that user's own connection.
4. Read tools: metabase_list_dashboards, metabase_get_dashboard, metabase_list_questions, metabase_run_question, metabase_list_databases, metabase_run_query. Write tools: metabase_create_question. Name "metabase" in the tools of a request; the model decides which tool to call.
5. Stay within the project's settings. If a tool is turned off or a call is refused, tell the user. Do not work around it.
6. Before a write, confirm with the user, unless they asked for that exact change.
7. For setup with a specific framework, follow its guide:
   https://www.agent200.com/connectors/metabase/framework/claude/
   https://www.agent200.com/connectors/metabase/framework/chatgpt/
   https://www.agent200.com/connectors/metabase/framework/claude-code/
   https://www.agent200.com/connectors/metabase/framework/muse/
   https://www.agent200.com/connectors/metabase/framework/grok-bot/
   https://www.agent200.com/connectors/metabase/framework/openai-dots/
   https://www.agent200.com/connectors/metabase/framework/claude-duet/
   https://www.agent200.com/connectors/metabase/framework/hermes-agent/
   https://www.agent200.com/connectors/metabase/framework/openclaw/

The code on the connector page is a conceptual example of the SDK. SDK reference: https://docs.igpt.ai/

This page is also available as Markdown, for agents and LLMs: /connectors/metabase.md

Guides

Use Metabase with the AI you build on.

Step-by-step setup for Metabase with a specific model, assistant or agent framework.

FAQ

Metabase and Agent200, answered.

Who handles the Metabase OAuth flow and the tokens?

Agent200. It runs the authorization, handles the callback on its own domain, stores the tokens and refreshes them. You do not build or maintain an OAuth integration for Metabase.

Does my agent act as the end user?

Within what that user authorized, yes. Every request names an end user, and Agent200 uses only that user's Metabase connection, never another user's. It cannot reach anything the user or Metabase has not authorized.

Can I limit what my agent can do in Metabase?

Yes. You choose which services, endpoints and authorization scopes are allowed in your Agent200 configuration, and the agent works only within them.

  • Tavily

    Tavily

    Web search built for agents, reached with your Agent200 API key.

    Agent200 API key Search and data
  • Exa

    Exa

    Search the web by meaning, read page contents and find similar pages, reached with your Agent200 API key.

    Agent200 API key Search and data
  • Firecrawl

    Firecrawl

    Scrape pages into Markdown, crawl and map sites, and extract structured data, reached with your Agent200 API key.

    Agent200 API key Search and data
  • Brave Search

    Brave Search

    Search the web, news, images and videos from Brave's independent index, reached with your Agent200 API key.

    Agent200 API key Search and data
200 OK

Start with Metabase.
Your agent does the rest.

Your users connect their own Metabase accounts. Agent200 handles the OAuth, the tokens and every tool call.

Book a demo.
See it in action.

Tell us who you are, then pick a time for a 1:1 session with an expert from our team.

We use your details only to email you about Agent200. See the Privacy Policy.

Pick a time. We'll take it from there.

A 1:1 session with an expert from our team, about what you are building.

Open in Calendly (opens in a new tab)