superset-mcp
superset-mcp
Apache Superset用のMCP (Model Context Protocol) サーバーです。AIアシスタントに、Supersetインスタンス(ダッシュボード、チャート、データセット、データベース、SQL実行)へのフルアクセス権を与えます。
前提条件
uv — Pythonパッケージマネージャー
実行中のApache Supersetインスタンスへのアクセス権
Related MCP server: Zaturn
セットアップ (所要時間: 約1分)
1. リポジトリのクローン
git clone <repo-url>
cd superset-mcp2. 認証情報の設定
cp .env.example .env.env を編集します:
SUPERSET_URL=http://your-superset-host:8088
SUPERSET_USERNAME=your-username
SUPERSET_PASSWORD=your-password3. 動作確認
uv run superset-mcpMCPサーバーがエラーなく起動することを確認してください。停止するには Ctrl+C を押します。
AIクライアントへの接続
/ABSOLUTE/PATH/TO/superset-mcp を、リポジトリをクローンした実際のパスに置き換えてください。
Claude Code
~/.claude/settings.json (またはプロジェクトレベルの .claude/settings.json) に追加します:
{
"mcpServers": {
"superset": {
"command": "uv",
"args": ["--directory", "/ABSOLUTE/PATH/TO/superset-mcp", "run", "superset-mcp"],
"env": {
"SUPERSET_URL": "http://your-superset-host:8088",
"SUPERSET_USERNAME": "your-username",
"SUPERSET_PASSWORD": "your-password"
}
}
}
}またはターミナルで実行します:
claude mcp add superset -- uv --directory /ABSOLUTE/PATH/TO/superset-mcp run superset-mcpClaude Desktop
~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
または %APPDATA%\Claude\claude_desktop_config.json (Windows) に追加します:
{
"mcpServers": {
"superset": {
"command": "uv",
"args": ["--directory", "/ABSOLUTE/PATH/TO/superset-mcp", "run", "superset-mcp"],
"env": {
"SUPERSET_URL": "http://your-superset-host:8088",
"SUPERSET_USERNAME": "your-username",
"SUPERSET_PASSWORD": "your-password"
}
}
}
}Cursor
プロジェクト内の .cursor/mcp.json またはグローバルの ~/.cursor/mcp.json に追加します:
{
"mcpServers": {
"superset": {
"command": "uv",
"args": ["--directory", "/ABSOLUTE/PATH/TO/superset-mcp", "run", "superset-mcp"],
"env": {
"SUPERSET_URL": "http://your-superset-host:8088",
"SUPERSET_USERNAME": "your-username",
"SUPERSET_PASSWORD": "your-password"
}
}
}
}利用可能なツール
ダッシュボード
ツール | 説明 |
| すべてのダッシュボードを一覧表示 (検索オプションあり) |
| IDまたはスラッグでダッシュボードを取得 |
| ダッシュボード内のすべてのチャートを取得 |
| ダッシュボードで使用されているすべてのデータセットを取得 |
| 新しいダッシュボードを作成 |
| タイトル、レイアウト、メタデータを更新 |
| ドラフトのダッシュボードを公開 |
| ダッシュボードを複製 |
| ダッシュボードを削除 |
チャート
ツール | 説明 |
| すべてのチャートを一覧表示 (検索オプションあり) |
| IDまたはUUIDでチャートを取得 |
| チャートの最新データを取得 |
| 新しいチャートを作成 |
| チャートのプロパティを更新 |
| チャートを削除 |
データセット
ツール | 説明 |
| すべてのデータセットを一覧表示 |
| IDまたはUUIDでデータセットを取得 |
| テーブルまたはSQLクエリから作成 |
| 既存の取得、または新規作成 |
| ソーステーブルからカラムを同期 |
| データセットのプロパティを更新 |
| 使用しているチャート/ダッシュボードを確認 |
| データセットを削除 |
データベース
ツール | 説明 |
| すべてのデータベース接続を一覧表示 |
| IDでデータベース接続を取得 |
| データベース内のスキーマを一覧表示 |
| スキーマ内のテーブルを一覧表示 |
| テーブルのカラム情報を取得 |
| テーブルの SELECT * テンプレートを取得 |
SQL
ツール | 説明 |
| SQLクエリを実行して結果を取得 |
| SQLクエリを整形 |
| クエリコストを見積もる (サポートされている場合) |
| 保存されたSQLクエリを一覧表示 |
| SQLクエリを保存 |
プロンプト例
AIアシスタントに接続後、以下のように指示できます:
"公開されているすべてのダッシュボードをリストアップして"
"'Sales Overview'という新しいダッシュボードを作成して"
"ダッシュボード11に含まれるすべてのチャートを見せて"
"データベース2でSQLクエリを実行: SELECT count() FROM orders"*
"'Revenue'ダッシュボードではどのデータセットが使われている?"
"データセット5を使用して、地域別の売上を示す棒グラフを作成して"
Available Tools
34 toolscopy_dashboardC
Create a copy of a dashboard.
Inputs: id_or_slug (required). Returns: object
create_chartC
Create a new chart.
Inputs: slice_name (required), viz_type (required), datasource_id (required), datasource_type, +3 more. Returns: object
create_dashboardC
Create a new dashboard.
Inputs: title (required), slug, published, json_metadata, +1 more. Returns: object
create_datasetB
Create a dataset from a table or SQL query.
Inputs: database_id (required), table_name (required), schema, sql, +1 more. Returns: object
delete_chartC
Delete a chart by ID.
Inputs: pk (required). Returns: object
delete_dashboardC
Delete a dashboard by ID.
Inputs: pk (required). Returns: object
delete_datasetB
Delete a dataset by ID.
Inputs: pk (required). Returns: object
estimate_query_costC
Estimate the cost of running a SQL query (supported databases only).
Inputs: database_id (required), sql (required), schema. Returns: object
execute_sqlB
Execute a SQL query and return results.
Inputs: database_id (required), sql (required), schema, limit. Returns: object
format_sqlB
Format/pretty-print a SQL query.
Inputs: sql (required). Returns: object
get_chartC
Get a chart by ID or UUID.
Inputs: id_or_uuid (required). Returns: object
get_chart_dataC
Fetch the latest data for a chart.
Inputs: pk (required). Returns: object
get_dashboardB
Get a single dashboard by ID or slug.
Inputs: id_or_slug (required). Returns: object
get_dashboard_chartsB
Get all charts in a dashboard.
Inputs: id_or_slug (required). Returns: object
get_dashboard_datasetsB
Get all datasets used by a dashboard.
Inputs: id_or_slug (required). Returns: object
get_databaseB
Get a database connection by ID.
Inputs: pk (required). Returns: object
get_datasetB
Get a dataset by ID or UUID.
Inputs: id_or_uuid (required). Returns: object
get_or_create_datasetB
Get an existing dataset or create it if it doesn't exist.
Inputs: database_id (required), table_name (required), schema. Returns: object
get_select_starC
Get a SELECT * query for a table (with schema).
Inputs: database_id (required), table_name (required), schema. Returns: object
get_table_metadataC
Get column metadata for a table.
Inputs: database_id (required), table_name (required), schema. Returns: object
list_chartsC
List charts. Optionally filter by name.
Inputs: page, page_size, search. Returns: object
list_dashboardsB
List dashboards. Optionally filter by title with search.
Inputs: page, page_size, search. Returns: object
list_databasesB
List all database connections configured in Superset.
Inputs: page, page_size. Returns: object
list_datasetsC
List datasets (virtual or physical tables).
Inputs: page, page_size, search. Returns: object
list_saved_queriesB
List saved SQL queries.
Inputs: page, page_size. Returns: object
list_schemasB
List all schemas in a database.
Inputs: database_id (required). Returns: object
list_tablesB
List tables in a database schema.
Inputs: database_id (required), schema. Returns: object
publish_dashboardC
Set a dashboard as published.
Inputs: pk (required). Returns: object
refresh_datasetC
Refresh the dataset's columns and metrics from the source.
Inputs: pk (required). Returns: object
save_queryC
Save a SQL query.
Inputs: label (required), sql (required), database_id (required), schema, +1 more. Returns: object
update_chartC
Update a chart's properties.
Inputs: pk (required), slice_name, description, viz_type, +2 more. Returns: object
update_dashboardC
Update an existing dashboard.
Inputs: pk (required), title, published, json_metadata, +1 more. Returns: object
update_datasetC
Update a dataset's properties.
Inputs: pk (required), table_name, description, sql, +3 more. Returns: object
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 34 tools
Every tool has a clearly distinct purpose targeting specific resources (charts, dashboards, datasets, databases, queries) and actions (create, get, list, update, delete, execute, format, etc.), with no apparent overlap or ambiguity. For example, get_chart, get_chart_data, and list_charts serve different functions without confusion.
Tool names follow a highly consistent verb_noun pattern throughout, such as create_chart, delete_dashboard, list_datasets, and update_dataset, with no deviations in style or convention. This predictability makes the tool set easy to navigate and understand.
With 34 tools, the count is borderline high for typical MCP server scopes, which often range from 3-15 tools. While it covers a comprehensive Superset domain, it may feel heavy and potentially overwhelming for agents, though not extreme like 50+ tools.
The tool surface provides complete CRUD/lifecycle coverage for charts, dashboards, datasets, and queries, along with execution, formatting, and metadata operations. There are no obvious gaps; agents can perform all core workflows from creation to deletion and analysis without dead ends.
Maintenance
Related MCP Connectors
MCP server that lets AI assistants use all OneSchema features exposed via the public API.
MCP server that lets your AI assistant build and deploy apps, hosting each one at a live URL.
MCP server for building and testing AI agents with multi-model experimentation and insights.
Draxlr's remote MCP server connects AI assistants to your SQL databases and dashboards. Explore schemas, run read-only queries, manage saved queries and dashboards, and export results, all with row-level security so each user sees only their own data.
Related MCP Servers
- AlicenseBqualityDmaintenanceMCP server enabling AI agents to programmatically connect to and control Apache Superset instances, allowing users to manage dashboards, charts, databases, datasets, and run SQL queries through natural language interactions.60195MIT
- AlicenseAqualityDmaintenanceAn open-source MCP server that connects to various data sources (SQL databases, CSV, Parquet files), allowing AI models to execute SQL queries and generate data visualizations for analytics and business intelligence.1275MIT
- AlicenseNot gradedqualityDmaintenanceAn MCP server for Apache Superset that enables AI models to manage dashboards, create charts with auto-detection, and execute SQL queries programmatically. It utilizes stateless HTTP transport to ensure compatibility with integration tools like Open WebUI.MIT
- AlicenseAqualityDmaintenanceAn MCP server that enables AI assistants to query databases, execute SQL, and manage Metabase resources like dashboards, cards, and collections through natural language.22MIT