Installation of the MCP Server
During SQLBits 2026, I had the opportunity to present on ‚The Good, the Bad, and the Ugly of Power BI Agentic Development.‘ Now, I want to take everything I’ve learned along the way and turn it into a dedicated blog series to help you make your first steps into Agentic Development.
A quick but important word of caution before we begin: Using these MCP servers means your data might be shared with outside services, such as third-party LLMs that aren’t covered by Microsoft’s compliance boundaries. Please remember that you are solely responsible for making sure your setup meets all of your company’s security, regulatory, and legal requirements.
Setup
I used the following setup:
Installation / Configuration
A prerequisite for installing the Power BI MCP Server is having Visual Studio Code installed. I won’t explain how to install Visual Studio Code. If you want to install der Power BI Modeling MCP Server you just need to press Install.
The installation itself is pretty fast.
You should define a working folder, where you will do your Power BI work. This can be a OneDrive Folder, Local Folder or something similar. In this case I’ve chosen a Local Folder.
Another important part is to have a mcp.json in your folder. Here is a simple example of an mcp.json for only Power BI:
MCP configuration (for manual registration):
{
"servers": {
"powerbi-modeling-mcp": {
"command": "C:\\MCPServers\\PowerBIModelingMCP\\extension\\server\\powerbi-modeling-mcp.exe",
"args": [
"--start"
],
"env": {}
}
}
}This is how the final result can look in Visual Studio Code.
As a best practice, I highly recommend keeping the Power BI file you want to work with right in your main project folder. For this series, I’ll be using the ‚Lion Dashboard’—a custom report I built specifically to experiment and play around with Agentic Development.
I copied a pbip Report in my Folder. This report has csv as a source. You can open the chat window for Github Copilot via Ctrl+Alt+I.
Connection to Power BI
A way to connect to Power BI is to have Power BI opened and use the command connect ‘Power BI Report’ in Power BI Desktop.
Here are some examples you can do?
Have fun and start exploring Power BI Agentic Development! In my next blog post, I’ll dive deeper into of more use cases and I try to explain how Agentic Development actually works.
Links
If you want to dive right in, here are a few essential links for working with Power BI Agentic Development:
Power BI Modeling MCP Server: The starting point to download the actual Power BI MCP Server. It explains the basic installation and setup. It gives you also information about governance and compliance. This was my starter for my journey and most text is coming from there.
Rui Romano’s Power BI Agentic Plugins: An incredible, hands-on GitHub repository. It provides pre-built „plugins“ (skills, agents, and connectors) designed to turn GitHub Copilot into a specialized developer for Power BI and Microsoft Fabric.
Data-Goblin’s Power BI Agentic Development Marketplace: A highly useful plugins by the one and only Data Goblin Kurt Buhler. It features a collection of plugins, skills, and hooks (like Best Practice Analyzer rules and DAX validation) specifically focused on improving how coding agents interact with Power BI.
This article is mainly based on: microsoft/powerbi-modeling-mcp: The Power BI Modeling MCP Server, brings Power BI semantic modeling capabilities to your AI agents.
You’re also welcome to watch the video from the last AI Slam for Finance and Controlling, where I discuss the future direction of Power BI development: KI Slam für Finance und Controlling Viktor Mendel
Screenshots: „Used with permission from Microsoft.“