# Metabase MCP for AI Agents AI Agent Connect

> Metabase MCP for AI Agents. It connects your Metabase instance to your AI client so you can search dashboards, audit SQL logic, and explore your BI collections using plain English.

## Overview
- **Category:** brain-trust
- **Price:** Free
- **Endpoint:** https://edge.vinkius.com/vk_preview_81usxhuffmH5qVmHtpmbFZWoVXrKKHjT3u2k9g8R/ai-agent-connect
- **Tags:** data-visualization, sql-queries, dashboarding, data-exploration, business-insights, open-source

## Description

Metabase MCP connects your business intelligence platform to your AI agent so you can manage reporting panels and audit data logic through natural conversation. You can pull your entire reporting structure into your chat to stop clicking through dozens of nested folders and sub-folders. It lets you see the underlying SQL logic of your cards, check dashboard layouts, and inventory your database connections without ever leaving your chat window. This is the fix for when you're tired of hunting for that one "Revenue" dashboard or trying to remember which database a specific report is actually hitting. You'll find it much easier to manage your reporting taxonomy because you can browse collections and see how everything is organized. By adding this to your Vinkius catalog, you turn your static dashboards into a queryable knowledge base that your agent can actually use to answer questions. It makes exploring your data warehouse connections and raw table mappings a simple chat command rather than a manual search in the settings menu. Instead of manually copying and pasting JSON representations of dashboard layouts to share with your team, you can just have your agent fetch the matrix for you. It simplifies how you audit your metrics by bringing the query logic straight to your workspace. You get a bird's-eye view of your entire BI environment, making it easy to spot gaps in your reporting or identify which cards need a logic refresh.

## Tools

### list_databases
list_databases lists the physical Database integrations. Use it to see which raw data sources are currently parsed by your platform.

### search_content
search_content executes a global search for reporting entities. Use it to find Dashboards, Cards, or tables using specific keywords.

### list_dashboards
list_dashboards lists all dashboards mapped in the platform. Use it to get a full inventory of your reporting panels.

### get_dashboard
get_dashboard retrieves the layout matrices for a dashboard. Use it to see how specific cards and filters are arranged.

### list_cards
list_cards returns a heavy list of raw questions. Use it to audit every card parsed inside your Metabase environment.

### get_card
get_card fetches the explicit mapping logic behind a specific BI Card. Use it to see the exact SQL or visualization rules driving a metric.

### list_collections
list_collections lists the structural BI Collections in your instance. Use it to understand how your reporting folders are organized.

## Prompt Examples

**Prompt:** 
```
List the last 5 dashboards created in my Metabase instance
```

**Response:** 
```
I've pulled the 5 newest dashboards for you. Here they are:

| Dashboard Name | ID |
| :--- | :--- |
| Q1 Sales Overview | 123 |
| User Engagement | 456 |
| Inventory Health | 789 |
| Revenue Tracker | 101 |
| Marketing Spend | 202 |

Would you like to see the cards attached to the sales overview?
```

**Prompt:** 
```
Search for any Metabase content related to 'Revenue'
```

**Response:** 
```
I found several matches in your Metabase instance:

* **Dashboard**: Annual Revenue Breakdown (ID: 987)
* **Card**: Monthly Recurring Revenue (ID: 654)
* **Table**: `public.revenue_events` 

Which one should I inspect further for you?
```

**Prompt:** 
```
Show me the configuration for card ID '654'
```

**Response:** 
```
Here is the configuration for the **Monthly Recurring Revenue** card (ID: 654):

* **Visualization**: Line Chart
* **Database**: `stripe_data`
* **Logic**: Uses a filtered aggregation on the `invoices` table.

Would you like me to pull the full structural JSON for this card?
```

## Capabilities

### Audit SQL logic
Get the raw queries behind your cards instantly.

### Find dashboards by keyword
Search your entire Metabase instance for specific reports.

### Browse folder structures
Navigate through BI collections to see how data is organized.

### View dashboard layouts
See the exact matrix and card mapping for any dashboard.

### Inventory database connections
List the physical data sources linked to your platform.

## Use Cases

### Auditing a revenue metric
A data analyst suspects a revenue number is wrong. They ask their agent to check the SQL for the 'Monthly Recurring Revenue' card using get_card.

### Finding a specific report
A marketing manager needs the 'Q4 Campaign Results' but can't find it. They ask their agent to search for 'Q4' using search_content.

### Mapping out the BI structure
A new team member wants to see how reports are organized. They ask their agent to list all collections to understand the reporting taxonomy.

### Checking database sources
An engineer needs to know which raw tables are available. They ask their agent to list the databases to see the parsed warehouse connections.

## Benefits

- Stop digging through nested folders with list_collections to find your reports.
- Audit SQL logic instantly with get_card without opening the Metabase UI.
- Find specific reports fast using search_content instead of manual browsing.
- Get a full inventory of your reporting panels with list_dashboards and list_cards.
- Understand your data architecture with list_databases to see your raw sources.
- Get exact dashboard layouts with get_dashboard to see card mappings clearly.

## How It Works

The bottom line is you get a searchable, queryable version of your Metabase environment inside your AI chat.

1. Subscribe to the Connector and provide your Metabase Base URL and API Key.
2. Connect your preferred AI client to the Vinkius host.
3. Ask your agent to find a specific report or explain a card's logic.

## Frequently Asked Questions

**Can I use the Metabase MCP to find specific reports?**
Yes, you can use the search tool to find any reports, dashboards, or tables across your entire Metabase instance using plain English keywords.

**How does the Metabase MCP help with SQL auditing?**
It allows your agent to pull the exact SQL logic and visualization rules behind any specific card so you can verify how metrics are calculated.

**Can I see my database connections with the Metabase MCP?**
Yes, you can ask your agent to list the physical database integrations that your Metabase platform is currently connected to.

**Will the Metabase MCP help me organize my BI folders?**
It lets you browse and list all your BI collections, making it much easier to understand and navigate your reporting taxonomy.

**Is there a way to see dashboard layouts using the Metabase MCP?**
Yes, you can request the layout matrix for any dashboard to see exactly how cards and filters are arranged in the UI.

**Can I search for keywords across all my Metabase dashboards?**
Yes, the search tool scans your entire Metabase instance for Dashboards, Cards, and tables that match your specific keywords.

**Does this work with self-hosted Metabase instances?**
Yes. You just need to provide your full Metabase Base URL (e.g., http://metabase.your-domain.com). Your agent will connect securely to your private instance using your API Key.

**Can I see the SQL code behind a specific Metabase card through my agent?**
Yes. Use the `get_card` tool with a specific Card ID. Your agent will retrieve the complete mapping logic, exposing the underlying SQL query (if it's a SQL question) or the structural query parameters used to generate the visualization.

**Can my agent list all databases currently connected to Metabase?**
Absolutely. Use the `list_databases` tool to identify all physical data warehouse integrations. Your agent will report the database names, types (e.g., PostgreSQL, Snowflake), and internal IDs available for querying.