mcp-metabase-server P

Metabase MCP Server

A comprehensive Model Context Protocol (MCP) server for Metabase integration, designed for enterprises and developers who want to supercharge their AI assistants with live analytics and business intelligence.

Developed and maintained by EaseCloud Technologies — your trusted partner for cloud-native, AI-driven, and data infrastructure solutions.

This is a TypeScript-based MCP server that implements full integration with Metabase API. It allows AI assistants to interact with Metabase, providing comprehensive access to:

• Complete CRUD Operations: Full lifecycle management for all Metabase entities
• Advanced Dashboard Management: Card positioning, parameters, subscriptions, and alerts
• User & Permission Management: Complete user lifecycle and permission system control
• Database Management: Connection management, schema sync, and metadata exploration
• Analytics & Monitoring: Usage statistics, activity tracking, and system health monitoring
• Advanced Query Features: Export, bookmarking, public sharing, and complex query execution
• Search & Discovery: Global search, popular items, and content recommendations
• Settings & Configuration: System settings and instance configuration management

70+ Tools Available covering all major Metabase functionality for enterprise usage.

Supported Metabase Versions

This MCP server is compatible with:

• Metabase v0.46.x and above (recommended: v0.48.x or later)
• Metabase Cloud (fully supported)
• Self-hosted Metabase instances (Docker, JAR, or cloud deployments)

Note: Some advanced features may require newer Metabase versions. For optimal compatibility, we recommend using Metabase v0.48.0 or later.

Why We Built This

At EaseCloud, we help companies modernize their data platforms and unlock the power of AI.
Metabase is a leading open-source BI tool — but connecting it with AI assistants like Claude, Cursor, and Windsurf requires complex API work.

This MCP server removes that barrier by providing 70+ ready-to-use tools that cover everything from dashboards to queries, permissions to monitoring.

Our goal is to make enterprise analytics accessible in natural language while showcasing our expertise in integration, automation, and cloud solutions.

Features

Resources

• Access Metabase resources via metabase:// URIs
• Dashboards: metabase://dashboard/{id} - Access dashboard details
• Cards/Questions: metabase://card/{id} - Access question/card details
• Databases: metabase://database/{id} - Access database information
• Collections: metabase://collection/{id} - Access collection details
• Users: metabase://user/{id} - Access user information
• Tables: metabase://table/{id} - Access table metadata
• Fields: metabase://field/{id} - Access field information
• JSON content type for structured data access

Core Data Management Tools

Dashboard Management

• list_dashboards - List all dashboards in Metabase
• create_dashboard - Create a new dashboard
• update_dashboard - Update an existing dashboard
• delete_dashboard - Delete/archive a dashboard
• get_dashboard_cards - Get all cards in a dashboard

Card/Question Management

• list_cards - List all questions/cards in Metabase
• create_card - Create a new question/card
• update_card - Update an existing question/card
• delete_card - Delete/archive a question/card
• execute_card - Execute a card and get results

Database Operations

• list_databases - List all databases in Metabase
• execute_query - Execute a SQL query against a database

Collections Management

• list_collections - List all collections
• create_collection - Create a new collection
• update_collection - Update an existing collection
• delete_collection - Delete a collection
• get_collection_items - Get all items in a collection
• move_to_collection - Move items between collections

User & Permission Management

• list_users - List all users
• create_user - Create a new user
• update_user - Update user details
• delete_user - Deactivate a user
• list_permissions - List permission groups and permissions
• list_permission_groups - List all permission groups
• create_permission_group - Create a new permission group
• update_permission_group - Update a permission group
• delete_permission_group - Delete a permission group
• add_user_to_group - Add user to permission group
• remove_user_from_group - Remove user from permission group

Advanced Query Features

• export_card_results - Export card results to CSV/JSON/Excel
• export_dashboard_data - Export dashboard data to various formats
• bookmark_card - Bookmark a card
• unbookmark_card - Remove card bookmark
• bookmark_dashboard - Bookmark a dashboard
• unbookmark_dashboard - Remove dashboard bookmark
• list_bookmarked_items - List all bookmarked items
• create_public_link - Create public sharing link
• disable_public_link - Disable public sharing
• get_public_link - Get public link information

Database Schema Operations

• get_database_schema - Get database schema information
• get_database_tables - Get all tables in a database
• get_table_metadata - Get table metadata
• get_table_fields - Get all fields in a table
• get_field_values - Get field values
• get_field_summary - Get field summary statistics
• analyze_field_values - Analyze field values

Search & Discovery

• search_content - Search across all Metabase content
• get_recent_items - Get recently viewed items
• get_popular_items - Get most popular cards and dashboards

Advanced Dashboard Features

• add_card_to_dashboard - Add a card to a dashboard with positioning
• remove_card_from_dashboard - Remove a card from a dashboard
• update_dashboard_card - Update card position, size, and settings
• get_dashboard_parameters - Get dashboard parameters
• update_dashboard_parameters - Update dashboard parameters
• create_dashboard_subscription - Create dashboard subscriptions/alerts
• list_dashboard_subscriptions - List all dashboard subscriptions
• update_dashboard_subscription - Update subscription settings
• delete_dashboard_subscription - Delete a dashboard subscription

Analytics & Monitoring

• get_user_activity - Get user activity and usage statistics
• get_content_usage - Get usage statistics for cards and dashboards
• get_system_usage_stats - Get system-wide usage statistics
• get_system_health - Get system health status

Database Connection Management

• create_database_connection - Create new database connections
• test_database_connection - Test database connectivity
• sync_database_schema - Sync database schema metadata
• get_database_sync_status - Get database schema sync status

Settings & Configuration

• get_metabase_settings - Get Metabase instance settings
• update_metabase_settings - Update instance settings

Configuration

Before running the server, you need to set environment variables for authentication. The server supports two methods:

1. API Key (Preferred):
   • METABASE_URL: The URL of your Metabase instance (e.g., https://your-metabase-instance.com).
   • METABASE_API_KEY: Your Metabase API key.
2. Username/Password (Fallback):
   • METABASE_URL: The URL of your Metabase instance.
   • METABASE_USERNAME: Your Metabase username.
   • METABASE_PASSWORD: Your Metabase password.

The server will first check for METABASE_API_KEY. If it's set, API key authentication will be used. If METABASE_API_KEY is not set, the server will fall back to using METABASE_USERNAME and METABASE_PASSWORD. You must provide credentials for at least one of these methods.

Example setup:

Using API Key:

Or, using Username/Password:

You can set these environment variables in your shell profile or use a .env file with a package like dotenv.

Development

Install dependencies:

Build the server:

For development with auto-rebuild:

Installation & Usage

Method 1: Using npx (Recommended)

The easiest way to run the server is using npx:

Method 2: Using Node.js directly

If you have the package installed locally or globally:

Or run from the built project:

Method 3: Using Docker

You can run the server using Docker:

Or using docker-compose:

Then run:

Integration with Claude Desktop

To use with Claude Desktop, add the server config:

On MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json On Windows: %APPDATA%/Claude/claude_desktop_config.json

Using npx:

Using Node.js directly:

Alternative authentication (username/password):

Environment Variables

The server supports the following environment variables:

• METABASE_URL (required): The URL of your Metabase instance
• METABASE_API_KEY (preferred): Your Metabase API key
• METABASE_USERNAME: Your Metabase username (fallback if API key not provided)
• METABASE_PASSWORD: Your Metabase password (fallback if API key not provided)

Debugging

Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector, which is available as a package script:

The Inspector will provide a URL to access debugging tools in your browser.

Testing

After configuring the environment variables as described in the "Configuration" section, you can manually test the server's authentication. The MCP Inspector (npm run inspector) is a useful tool for sending requests to the server.

1. Testing with API Key Authentication

1. Set the METABASE_URL and METABASE_API_KEY environment variables with your Metabase instance URL and a valid API key.
2. Ensure METABASE_USERNAME and METABASE_PASSWORD are unset or leave them, as the API key should take precedence.
3. Start the server: npm run build && node build/index.js (or use your chosen method for running the server, like via Claude Desktop config).
4. Check the server logs. You should see a message indicating that it's using API key authentication (e.g., "Using Metabase API Key for authentication.").
5. Using an MCP client or the MCP Inspector, try calling a tool, for example, tools/call with {"name": "list_dashboards"}.
6. Verify that the tool call is successful and you receive the expected data.

2. Testing with Username/Password Authentication (Fallback)

1. Ensure the METABASE_API_KEY environment variable is unset.
2. Set METABASE_URL, METABASE_USERNAME, and METABASE_PASSWORD with valid credentials for your Metabase instance.
3. Start the server.
4. Check the server logs. You should see a message indicating that it's using username/password authentication (e.g., "Using Metabase username/password for authentication." followed by "Authenticating with Metabase using username/password...").
5. Using an MCP client or the MCP Inspector, try calling the list_dashboards tool.
6. Verify that the tool call is successful.

3. Testing Authentication Failures

• Invalid API Key:
  1. Set METABASE_URL and an invalid METABASE_API_KEY. Ensure METABASE_USERNAME and METABASE_PASSWORD variables are unset.
  2. Start the server.
  3. Attempt to call a tool (e.g., list_dashboards). The tool call should fail, and the server logs might indicate an authentication error from Metabase (e.g., "Metabase API error: Invalid X-API-Key").
• Invalid Username/Password:
  1. Ensure METABASE_API_KEY is unset. Set METABASE_URL and invalid METABASE_USERNAME/METABASE_PASSWORD.
  2. Start the server.
  3. Attempt to call a tool. The tool call should fail due to failed session authentication. The server logs might show "Authentication failed" or "Failed to authenticate with Metabase".
• Missing Credentials:
  1. Unset METABASE_API_KEY, METABASE_USERNAME, and METABASE_PASSWORD. Set only METABASE_URL.
  2. Attempt to start the server.
  3. The server should fail to start and log an error message stating that authentication credentials (either API key or username/password) are required (e.g., "Either (METABASE_URL and METABASE_API_KEY) or (METABASE_URL, METABASE_USERNAME, and METABASE_PASSWORD) environment variables are required").

About EaseCloud Technologies

EaseCloud is a cloud consulting and solutions company specializing in:

• Cloud-native application development
• AI & automation integrations
• DevOps and infrastructure management
• Data analytics and BI platform consulting

We built this project to contribute to the open-source MCP ecosystem while demonstrating our deep expertise in data-driven decision systems.

👉 If your team is adopting Metabase at scale or looking to integrate AI with your BI stack, get in touch with us — we provide consulting, customization, and managed support for enterprises.

💡 Work with EaseCloud
Need help deploying, customizing, or scaling Metabase in production?
We provide end-to-end support for BI, AI, and cloud infrastructure.

📧 Contact us: support@easecloud.io
🌐 Learn more: https://easecloud.io

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

For support, please contact us at support@easecloud.io or visit our website at EaseCloud.io.

Bug Reports & Issues

Found a bug or have a feature request? We'd love to hear from you!

🐛 Report Issues: Create a bug report on GitHub

When reporting issues, please include:

• Metabase version you're using
• MCP server version
• Steps to reproduce the issue
• Expected vs actual behavior
• Any error messages or logs

Contributing

Contributions are welcome! Please visit our GitHub repository to submit issues or pull requests.