Analyze Powerdrill Data with Natural Language

An MCP server for Powerdrill datasets that lets Claude analyze data and execute queries using natural-language questions, including listing datasets and data sources, creating new datasets, uploading local files as data sources, and running natural-language analysis jobs grouped into sessions.

Use this when Claude needs to list or query Powerdrill datasets, create a new dataset or upload a file as a data source, or run a natural-language analysis job against existing data - for example "What datasets are available in my Powerdrill account?" or "Create a new dataset called 'Sales Analytics'." It requires an active Powerdrill Team account with a User ID and Project API Key - it is not usable without a Powerdrill subscription.

• mcp_powerdrill_list_datasets: lists available datasets from your Powerdrill account.
• mcp_powerdrill_get_dataset_overview: retrieves detailed information about a specific dataset.
• mcp_powerdrill_create_job: creates a job to analyze data using natural-language questions.
• mcp_powerdrill_create_session: creates a new session for grouping related jobs.
• mcp_powerdrill_list_data_sources: lists data sources within a specific dataset.
• mcp_powerdrill_list_sessions: lists sessions from your Powerdrill account.
• mcp_powerdrill_create_dataset: creates a new dataset in your Powerdrill account.
• mcp_powerdrill_create_data_source_from_local_file: uploads a local file to a specified dataset as a new data source.

Via Smithery:

npx -y @smithery/cli install @powerdrillai/powerdrill-mcp --client claude

Or install globally (npm install -g @powerdrillai/powerdrill-mcp), run directly with npx @powerdrillai/powerdrill-mcp, or build from source and run the provided setup.sh. Configure Claude Desktop or Cursor with:

{
  "powerdrill": {
    "command": "npx",
    "args": [
      "-y",
      "@powerdrillai/powerdrill-mcp@latest"
    ],
    "env": {
      "POWERDRILL_USER_ID": "your_actual_user_id",
      "POWERDRILL_PROJECT_API_KEY": "your_actual_project_api_key"
    }
  }
}

Both POWERDRILL_USER_ID and POWERDRILL_PROJECT_API_KEY are required for authentication, and video tutorials are provided for setting up a Powerdrill Team account and its API keys.

Data teams already on a Powerdrill Team plan who want Claude or another MCP client to list, create, and query Powerdrill datasets using natural language instead of the Powerdrill web UI. Typical prompts include "Show me all my datasets", "Upload the file /Users/your_name/Downloads/sales_data.csv to dataset {dataset_id}", and "Tell me more about this dataset: {dataset_id}", and the server integrates directly with both Claude Desktop and other MCP-compatible clients such as Cursor, using the same NPX-based configuration in each case. Beyond the Node-based MCP server itself, the wider Powerdrill ecosystem also offers separate Node.js and Python client options, and the project notes require an active Powerdrill Team account rather than a free-tier account, since the underlying dataset and job APIs are gated behind that plan.