mcpField
P
Power Bi Mcp Server

Power Bi Mcp Server

Local
github.com/michaelmckinleyconsultingData & Analytics4Updated 49w ago

About Power Bi Mcp Server

An MCP server for users who want AI assistants to operate on Power BI workspaces, reports, datasets, and dashboards. It supports DAX queries, report exports, and push datasets, with OAuth2 authentication through Microsoft Entra ID.

Key features

  1. 01List workspaces and manage members
  2. 02Clone, export, and rebind reports
  3. 03Run DAX queries and refresh datasets
  4. 04Manage dataset refresh schedules
  5. 05View dashboards and tiles
  6. 06Create push datasets and add data

Use cases

  1. 01Query regional totals in a sales dataset
  2. 02Export a quarterly report as PDF or PPTX
  3. 03Refresh a dataset and check its history
  4. 04Add a user to an analytics workspace

MCP support

The MCP server exposes operations for Power BI workspaces, reports, datasets, and dashboards to compatible AI assistants. Authentication supports interactive sign-in and app credentials.

Configuration

Local server config is the same for every client. Copy it into your client’s MCP config file.

No install method listed yet

We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.

Repository

Tools

No tool list yet

Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.

Details

From the project README

Power BI MCP Server

A Model Context Protocol (MCP) server that enables AI assistants to interact with Power BI workspaces, datasets, reports, and dashboards programmatically.

🚀 Features

  • Workspace Management: List and manage Power BI workspaces
  • Report Operations: Access, clone, export, and rebind reports
  • Dataset Management: Execute DAX queries, refresh datasets, manage schedules
  • Dashboard Access: List and interact with dashboards
  • Push Datasets: Create and manage push datasets with real-time data
  • Authentication: Secure OAuth2 authentication with Microsoft Entra ID

📋 Prerequisites

  • Node.js (v18 or higher)
  • npm or yarn
  • Power BI Pro or Premium license
  • Microsoft Entra ID app registration
  • MCP-compatible AI assistant (Claude Desktop, etc.)

🔧 Installation

  1. Clone the repository:
git clone https://github.com/michaelmckinleyconsulting/powerbi-mcp-server.git
cd powerbi-mcp-server
  1. Install dependencies:
npm install
  1. Build the project:
npm run build

⚙️ Configuration

1. Microsoft Entra ID Setup
  1. Register an application in Azure Portal
  2. Configure API permissions for Power BI Service
  3. Note your:
    • Client ID
    • Tenant ID
    • Client Secret (if using app-only auth)
2. Environment Variables

Create a .env file in the root directory (see .env.example for template):

PBI_PUBLIC_CLIENT_ID=your_client_id
PBI_TENANT_ID=your_tenant_id  # Optional, defaults to 'common'
PBI_SCOPES=https://analysis.windows.net/powerbi/api/.default  # Optional, or specify custom scopes

Note: This server uses Authorization Code flow with PKCE (no client secret required). Configure your Azure app with redirect URIs:

  • http://localhost
  • http://127.0.0.1
3. MCP Configuration

Add to your MCP client configuration:

Claude Desktop
{
  "mcpServers": {
    "powerbi": {
      "command": "node",
      "args": ["path/to/powerbi-mcp-server/dist/index.js"],
      "env": {
        "PBI_PUBLIC_CLIENT_ID": "your_client_id",
        "PBI_TENANT_ID": "your_tenant_id"
      }
    }
  }
}
VS Code MCP Extension
{
  "mcp.servers": {
    "powerbi": {
      "command": "node",
      "args": ["path/to/powerbi-mcp-server/dist/index.js"],
      "transport": "stdio",
      "env": {
        "PBI_PUBLIC_CLIENT_ID": "your_client_id"
      }
    }
  }
}

📖 Usage

Basic Operations

Once configured, your AI assistant can:

  • List workspaces: "Show me all my Power BI workspaces"
  • Execute DAX: "Run a DAX query to get sales by region"
  • Export reports: "Export the quarterly report to PDF"
  • Refresh datasets: "Refresh the sales dataset"
  • Manage access: "Add [email protected] to the Analytics workspace"
Example Prompts
"List all reports in my Sales workspace"
"Execute this DAX query against the Finance dataset: EVALUATE SUMMARIZE(...)"
"Export the Monthly KPI report to PowerPoint"
"Show me the refresh history for the Customer dataset"
"Create a push dataset for real-time monitoring"

🔐 Authentication

The server supports two authentication methods:

  1. Interactive (Recommended): User signs in via browser
  2. App-Only: Uses client credentials (requires admin consent)

On first run, the server will prompt for authentication via your default browser.

📁 Project Structure

powerbi-mcp-server/
├── src/
│   ├── index.ts           # Main server entry point
│   ├── auth/              # Authentication logic
│   ├── handlers/          # Request handlers
│   ├── types/             # TypeScript definitions
│   └── utils/             # Utility functions
├── dist/                  # Compiled JavaScript (git-ignored)
├── package.json
├── tsconfig.json
└── README.md

🛠️ Development

Running in Development Mode
npm run dev
Running Tests
npm test
Building for Production
npm run build

📊 Supported Power BI Operations

Workspaces (Groups)
  • List workspaces
  • Get workspace users
  • Add/remove users
Reports
  • List reports
  • Get report metadata
  • Clone reports
  • Export to various formats (PDF, PPTX, PNG)
  • Rebind to different datasets
Datasets
  • List datasets
  • Execute DAX queries
  • Trigger refreshes
  • Manage refresh schedules
  • Get refresh history
Dashboards
  • List dashboards
  • Get dashboard tiles
Push Datasets
  • Create push datasets
  • Add rows to tables
  • Clear table data

🔧 Troubleshooting

Common Issues
  1. Authentication fails: Ensure your Azure app registration has proper Power BI permissions
  2. No workspaces found: Verify your account has access to Power BI workspaces
  3. DAX queries fail: Check dataset permissions and query syntax
  4. Export timeouts: Large reports may take time; the server handles async polling
Debug Mode

Enable verbose logging:

DEBUG=powerbi:* node dist/index.js

📄 License

This project is source-available under a Modified MIT License with Commons Clause - see the LICENSE file for details.

Non-Commercial Use: Free for personal, educational, and internal company use.

Commercial Use: Requires a separate commercial license. Contact [[email protected]] for commercial licensing.

This is not OSI-approved "open source" due to the commercial restrictions, but allows full community use for non-commercial purposes.

🤝 Contributing

We welcome contributions! Please note:

  1. By contributing, you agree that your contributions will be licensed under the same license
  2. For substantial changes, please open an issue first to discuss
  3. Ensure all tests pass and add tests for new features

💬 Support

🙏 Acknowledgments

  • Built on the Model Context Protocol specification
  • Uses Microsoft Power BI REST APIs
  • Inspired by the MCP community

Note: This is not an official Microsoft product. Power BI is a trademark of Microsoft Corporation.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.