powerbi-mcp

Overview

Power BI MCP Server bridges AI assistants with Microsoft Power BI through the Model Context Protocol. It provides a unified interface for querying data, managing models, and performing bulk operations across local .pbix files and cloud-hosted datasets. The server features advanced security with PII detection, audit logging, and configurable access policies. Its standout capability is the PBIP Connector, which enables safe refactoring operations that update both semantic models and report visuals simultaneously, preventing common issues where renaming tables breaks report references.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• desktop_discover_instances

  Auto-discover running Power BI Desktop instances

  Try:Find all running Power BI Desktop instancesTry:List available Power BI Desktop connections
• desktop_connect

  Connect to a specific Power BI Desktop instance with optional RLS role

  Try:Connect to my Power BI Desktop instanceTry:Connect to Power BI with Sales role
• desktop_execute_dax

  Execute DAX queries against Power BI Desktop model with security processing

  Try:Calculate total sales last yearTry:Show top 10 customers by revenue
• list_workspaces

  List accessible Power BI Service workspaces

  Try:Show my Power BI workspacesTry:List all available workspaces
• list_datasets

  List datasets in a specified Power BI workspace

  Try:List datasets in Sales workspaceTry:What datasets are in Finance workspace?
• execute_dax

  Execute DAX queries against Power BI Service datasets

  Try:Show monthly revenue trendsTry:Calculate product profit margin
• desktop_list_rls_roles

  List Row-Level Security roles defined in the model

  Try:Show all RLS rolesTry:List security roles in model
• desktop_set_rls_role

  Activate a specific RLS role for testing data access

  Try:Enable Sales role for testingTry:Set RLS role to Manager
• scan_table_dependencies

  Analyze dependencies before renaming tables/columns/measures

  Try:Check what depends on Sales tableTry:Show impact of renaming Customer column
• batch_update_measures

  Bulk update measure expressions in the model

  Try:Update all margin calculationsTry:Change all sales measures to use new category
• pbip_rename_tables

  Safely rename tables in PBIP projects (updates model + visuals + DAX)

  Try:Rename Customers table to AccountsTry:Change Orders to SalesOrders
• pbip_load_project

  Load a PBIP project for editing operations

  Try:Load Sales Report projectTry:Open Finance PBIP file

Comparable tools

Installation

# Install via pip
pip install powerbi-mcp

# Configure Claude Desktop
{
  "mcpServers": {
    "powerbi": {
      "command": "python",
      "args": ["-m", "powerbi_mcp"],
      "env": {
        "POWERBI_MCP_WORKSPACE": "your_workspace_path"
      }
    }
  }
}

FAQ

Can this server work with both Power BI Desktop and Power BI Service?

Yes, the server supports dual connectivity - it can interact with both local Power BI Desktop instances and cloud-hosted Power BI Service through XMLA endpoints.

How does the server handle renaming tables without breaking visuals?

The PBIP Connector uses text-based editing of TMDL files and JSON manipulation of report definitions to update both model layer and report layer references simultaneously, ensuring no visual references are broken after renaming operations.

Compare powerbi-mcp with