Metabase MCP Server
Enables AI assistants to access and manage Metabase analytics through 96 tools for dashboards, cards, databases, tables, collections, and query execution.
README
Metabase MCP Server
A Model Context Protocol (MCP) server for Metabase that gives AI assistants full access to your analytics platform — dashboards, cards, databases, tables, collections, and more.
Developed and maintained by EaseCloud — cloud-native, AI-driven, and data infrastructure solutions.
This is a fork of
@easecloudio/mcp-metabase-server, extended with a manual session-token authentication strategy (METABASE_SESSION_TOKEN) and query result row limiting (METABASE_MCP_MAX_ROWS). See Authentication and Row Limiting.
Quick Start
export METABASE_URL=https://your-metabase-instance.com
export METABASE_API_KEY=your_metabase_api_key
npx @easecloudio/mcp-metabase-server
96 Tools Available
| Domain | Tools |
|---|---|
| Dashboard Management | 27 |
| Card / Question Management | 21 |
| Database Management | 16 |
| Table Management | 17 |
| Collections, Users & Search | 13 |
| Schema Cache (SQL→MBQL) | 2 |
Supported Metabase Versions
- Metabase v0.46.x and above (recommended: v0.48.x or later)
- Metabase Cloud (fully supported)
- Self-hosted instances (Docker, JAR, or cloud deployments)
Installation
npx (Recommended)
npx @easecloudio/mcp-metabase-server
Global install
npm install -g @easecloudio/mcp-metabase-server
mcp-metabase-server
Docker
docker build -t mcp-metabase-server .
docker run -it --rm \
-e METABASE_URL=https://your-metabase-instance.com \
-e METABASE_API_KEY=your_metabase_api_key \
mcp-metabase-server
Configuration
Authentication
API Key (preferred):
METABASE_URL=https://your-metabase-instance.com
METABASE_API_KEY=your_metabase_api_key
Manual session token (for SSO-only instances):
If your Metabase instance only supports SSO login and you have no local password or self-service API key, copy the metabase.SESSION cookie value from an already-authenticated browser session (DevTools > Application > Cookies) and use it directly:
METABASE_URL=https://your-metabase-instance.com
METABASE_SESSION_TOKEN=your_metabase_session_cookie_value
This is a long-lived credential (typically valid up to Metabase's session timeout, ~14 days by default) with no way to refresh it automatically — treat it like a password. When it expires, the server raises a clear error telling you to copy a fresh cookie value and restart.
Username / Password (fallback):
METABASE_URL=https://your-metabase-instance.com
METABASE_USERNAME=your_username
METABASE_PASSWORD=your_password
Precedence when more than one credential is set: API key > session token > username/password. Pin a strategy explicitly with METABASE_AUTH_METHOD=api_key|session_token|password.
Copy .env.example to .env and fill in your values.
Row Limiting
METABASE_MCP_MAX_ROWS (default 2000) caps the number of rows returned by execute_query and execute_card to protect server memory and the LLM's context window. When a result is truncated, the response includes _truncated: true, _total_rows, and _rows_returned alongside the (capped) rows array.
METABASE_MCP_MAX_ROWS=2000
Tool Filtering (TOOL_MODE)
Set TOOL_MODE to control which tools are exposed to the AI. Useful for limiting surface area or preventing accidental writes.
| Mode | Description |
|---|---|
all |
Every available tool (default) |
essential |
Core read + execute tools only |
read |
All non-destructive tools |
write |
All tools including create / update / delete |
TOOL_MODE=essential # smallest surface area
TOOL_MODE=read # read-only
TOOL_MODE=all # everything (default)
Claude Desktop Integration
MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%/Claude/claude_desktop_config.json
Using npx:
{
"mcpServers": {
"metabase": {
"command": "npx",
"args": ["@easecloudio/mcp-metabase-server"],
"env": {
"METABASE_URL": "https://your-metabase-instance.com",
"METABASE_API_KEY": "your_metabase_api_key"
}
}
}
}
With tool filtering:
{
"mcpServers": {
"metabase": {
"command": "npx",
"args": ["@easecloudio/mcp-metabase-server"],
"env": {
"METABASE_URL": "https://your-metabase-instance.com",
"METABASE_API_KEY": "your_metabase_api_key",
"TOOL_MODE": "essential"
}
}
}
}
Using a local build:
{
"mcpServers": {
"metabase": {
"command": "node",
"args": ["/path/to/metabase-server/dist/index.js"],
"env": {
"METABASE_URL": "https://your-metabase-instance.com",
"METABASE_API_KEY": "your_metabase_api_key"
}
}
}
}
Username / password fallback:
{
"mcpServers": {
"metabase": {
"command": "npx",
"args": ["@easecloudio/mcp-metabase-server"],
"env": {
"METABASE_URL": "https://your-metabase-instance.com",
"METABASE_USERNAME": "your_username",
"METABASE_PASSWORD": "your_password"
}
}
}
}
Available Tools
<details> <summary><strong>Dashboard Management (27 tools)</strong></summary>
Core CRUD
| Tool | Description |
|---|---|
list_dashboards |
List all dashboards |
get_dashboard |
Get a specific dashboard by ID |
create_dashboard |
Create a new dashboard |
update_dashboard |
Update an existing dashboard |
delete_dashboard |
Delete / archive a dashboard |
copy_dashboard |
Duplicate a dashboard (shallow or deep copy) |
save_dashboard |
Save a complete dashboard object with nested data |
save_dashboard_to_collection |
Move a dashboard into a specific collection |
search_dashboards |
Search dashboards by name or description |
favorite_dashboard |
Mark a dashboard as a favourite |
unfavorite_dashboard |
Remove a dashboard from favourites |
Card Layout
| Tool | Description |
|---|---|
get_dashboard_cards |
Get all cards in a dashboard |
add_card_to_dashboard |
Add a card with positioning |
add_text_block |
Add a text or heading block to a dashboard |
remove_card_from_dashboard |
Remove a card |
update_dashboard_card |
Update card position, size, and settings |
update_dashboard_cards |
Bulk-replace all cards on a dashboard |
update_dashcard |
Update a specific dashcard's properties |
execute_dashboard_card |
Execute a specific dashcard and return results |
Public Sharing & Embedding
| Tool | Description |
|---|---|
create_dashboard_public_link |
Create a public sharing link |
delete_dashboard_public_link |
Remove the public link |
list_public_dashboards |
List dashboards with public links |
list_embeddable_dashboards |
List dashboards enabled for embedding |
Revision, Audit & Discovery
| Tool | Description |
|---|---|
get_dashboard_revisions |
Get revision history (audit trail) |
revert_dashboard |
Revert to a previous revision |
get_dashboard_related |
Get related content suggestions |
get_dashboard_queries |
Extract all card queries with resolved field IDs |
</details>
<details> <summary><strong>Card / Question Management (21 tools)</strong></summary>
Core CRUD
| Tool | Description |
|---|---|
list_cards |
List all questions / cards |
get_card |
Get a card by ID (includes full SQL / MBQL query) |
create_card |
Create a new question |
update_card |
Update an existing question |
delete_card |
Delete / archive a question |
copy_card |
Duplicate a card |
move_cards |
Move one or more cards to a different collection |
move_cards_to_collection |
Bulk-move cards from a source collection |
Execution & Export
| Tool | Description |
|---|---|
execute_card |
Run a card and return results |
execute_pivot_card_query |
Run a card formatted as a pivot table |
export_card_result |
Export results as CSV or JSON |
Parameters
| Tool | Description |
|---|---|
get_card_param_values |
Get available values for a card parameter |
search_card_param_values |
Search / filter parameter values |
get_card_param_remapping |
Get how parameter values are remapped for display |
Metadata & Discovery
| Tool | Description |
|---|---|
get_card_query_metadata |
Get column types and display names |
get_card_dashboards |
List dashboards containing this card |
get_card_series |
Get series data or related card suggestions |
Public Sharing & Embedding
| Tool | Description |
|---|---|
create_card_public_link |
Create a public sharing link |
delete_card_public_link |
Remove the public link |
list_public_cards |
List cards with public links |
list_embeddable_cards |
List cards enabled for embedding |
</details>
<details> <summary><strong>Database Management (16 tools)</strong></summary>
Core CRUD
| Tool | Description |
|---|---|
list_databases |
List all database connections |
get_database |
Get a specific database by ID |
create_database_connection |
Create a new database connection |
update_database |
Update a database connection's config or credentials |
delete_database |
Permanently remove a database connection |
add_sample_database |
Add the built-in H2 sample database with demo data |
Querying
| Tool | Description |
|---|---|
execute_query |
Execute a SQL query against a database |
Schema & Sync
| Tool | Description |
|---|---|
get_database_schema |
Get schema information |
get_database_tables |
Get all tables in a database |
get_database_metadata |
Get full metadata including all tables and field IDs |
list_database_schemas |
List all schemas within a database |
sync_database_schema |
Trigger a schema metadata sync |
get_database_sync_status |
Check sync status |
Diagnostics
| Tool | Description |
|---|---|
check_database_health |
Check connection health |
validate_database |
Validate connection settings before saving |
test_database_connection |
Test an existing connection |
</details>
<details> <summary><strong>Table Management (17 tools)</strong></summary>
Metadata
| Tool | Description |
|---|---|
list_tables |
List tables (optionally filtered by database) |
get_table |
Get table metadata by ID |
get_table_metadata |
Get full query metadata including field IDs and types |
get_table_fks |
Get foreign key relationships |
get_table_related |
Find related tables via FK relationships |
get_table_data |
Get a sample data preview from a table |
get_field_id |
Look up a field ID by table ID + column name (returns MBQL ref) |
Card Virtual Tables
| Tool | Description |
|---|---|
get_card_table_fks |
Get FK relationships for a card's virtual table |
get_card_table_query_metadata |
Get query metadata for a card's virtual table |
Management
| Tool | Description |
|---|---|
update_table |
Update display name, description, or visibility |
update_tables |
Bulk-update multiple tables with the same config |
reorder_table_fields |
Change the display order of fields |
sync_table_schema |
Trigger a schema sync for a specific table |
rescan_table_field_values |
Rescan field values (updates filter dropdowns) |
discard_table_field_values |
Discard cached field values |
CSV Upload (Metabase-managed tables)
| Tool | Description |
|---|---|
append_csv_to_table |
Append new rows from CSV content |
replace_table_csv |
Replace all table data with new CSV content |
</details>
<details> <summary><strong>Collections, Users & Search (13 tools)</strong></summary>
Collections
| Tool | Description |
|---|---|
list_collections |
List all collections |
create_collection |
Create a new collection |
get_collection |
Get collection details |
update_collection |
Update name, description, color, or parent |
delete_collection |
Delete a collection and its contents |
get_collection_items |
List cards, dashboards, and sub-collections |
move_to_collection |
Move a card or dashboard to a different collection |
Users
| Tool | Description |
|---|---|
list_users |
List all users |
create_user |
Create a new user |
Permissions
| Tool | Description |
|---|---|
list_permission_groups |
List all permission groups |
create_permission_group |
Create a new permission group |
Search
| Tool | Description |
|---|---|
search_content |
Search across all Metabase content |
</details>
<details> <summary><strong>Schema Cache — SQL to MBQL (2 tools)</strong></summary>
These tools enable an AI to convert native SQL questions into interactive MBQL questions. Metabase has no REST API for this conversion — it requires field IDs, which these tools cache locally.
How it works:
get_card— extract the SQL anddatabase_idget_schema_cache— get tables + field IDs for MBQL references like["field", 42, null]- AI translates SQL → MBQL using the cached field IDs
create_card— save the new interactive question
Cache is stored at ~/.easecloud/metabase-mcp/cache/{url-hash}/ with a 24-hour TTL and scoped per Metabase instance.
| Tool | Description |
|---|---|
get_schema_cache |
Return cached schema (auto-fetches if missing or stale) |
refresh_schema_cache |
Force-refresh cache for one or all databases |
</details>
MCP Resources
Access Metabase entities directly via metabase:// URIs:
| URI | Description |
|---|---|
metabase://dashboard/{id} |
Dashboard details |
metabase://card/{id} |
Card / question details |
metabase://database/{id} |
Database information |
metabase://collection/{id} |
Collection details |
metabase://user/{id} |
User information |
metabase://table/{id} |
Table metadata |
metabase://field/{id} |
Field information |
Development
npm install
npm run build # compile TypeScript → dist/
npm run watch # incremental rebuild
npm run dev # build + start
npm run inspector # launch MCP Inspector for debugging
Debugging
MCP servers communicate over stdio, which makes direct debugging awkward. Use the MCP Inspector:
npm run inspector
The Inspector provides a browser UI for sending tool calls and inspecting responses.
About EaseCloud
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 expertise in integration, automation, and cloud solutions.
If your team is adopting Metabase at scale or looking to integrate AI with your BI stack, get in touch — we provide consulting, customization, and managed support for enterprises.
📧 support@easecloud.io 🌐 easecloud.io
Bug Reports & Issues
Found a bug or have a feature request?
Please include:
- Metabase version
- MCP server version
- Steps to reproduce
- Expected vs actual behavior
- Any error messages or logs
Contributing
Contributions are welcome. Visit the GitHub repository to submit issues or pull requests.
License
MIT — see LICENSE for details.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。