Microsoft Fabric

You run a Microsoft Fabric MCP Server that lets AI assistants interact with Power BI and Fabric services. It provides capabilities to list and query datasets, run DAX queries, refresh data, manage workspaces, and handle Fabric notebooks and data uploads. This guide shows you how to install, configure, and use the server in practical steps.

How to use

After you start the MCP server, you can use your MCP client to perform common tasks such as listing accessible workspaces, listing Power BI datasets, executing DAX queries against a dataset, and triggering dataset refreshes. You can also create Fabric notebooks and upload data to a Fabric data warehouse.

Typical usage patterns include asking your AI assistant to: 1) show all workspaces you can access, 2) list datasets in a workspace, 3) run a DAX query on a specific dataset, 4) refresh a dataset, 5) create a new Fabric notebook, or 6) upload data to a data warehouse. Your MCP client will invoke the available tools to perform these actions against your configured Power BI and Fabric resources.

How to install

Prerequisites you need before installing are Node.js and npm on your system. Ensure you have a working Node.js environment by confirming node -v and npm -v return valid versions.

Step 1: Install the MCP server package globally so you can run it from anywhere.

npm install -g @strainprint/microsoft-fabric-mcp

Step 2: Prepare environment variables for authentication and workspace scope. Create a .env file or export the variables in your shell.

# Required
AZURE_TENANT_ID=your-tenant-id
AZURE_CLIENT_ID=your-client-id
AZURE_CLIENT_SECRET=your-client-secret

# Optional
POWERBI_WORKSPACE_ID=workspace-id  # If not set, uses workspace-agnostic endpoints

Additional sections

Configuration is done through environment variables that grant access to Azure, Power BI, and Fabric services. The server uses these credentials to authenticate and perform actions on your behalf. Make sure the service principal has the necessary permissions in Power BI and that tenant settings allow service principals to call Power BI APIs.

To run the server from the command line alongside a client configuration, you can use a local runtime approach that does not require a remote endpoint. The following example shows how to invoke the MCP runtime with the package name and arguments.

npx -y @strainprint/microsoft-fabric-mcp

Troubleshooting

403 Forbidden Errors often indicate permission issues. Ensure the service principal is added to the Power BI workspace and that API permissions are granted in the Azure portal. Verify tenant settings allow service principals to access Power BI APIs and that the principal is in the correct security group.

Token issues may arise if the client secret expires or the tenant ID is incorrect. Validate that AZURE_TENANT_ID, AZURE_CLIENT_ID, and AZURE_CLIENT_SECRET are correct and active.

Available tools

get_workspaces

Retrieves all Power BI and Fabric workspaces accessible to the configured identity.

get_powerbi_datasets

Lists all datasets within the configured workspace.

execute_dax_query

Executes a DAX query against a specified dataset and returns results.

refresh_dataset

Triggers a refresh of a specified Power BI dataset.

create_notebook

Creates a new Microsoft Fabric notebook within a workspace.

upload_to_datawarehouse

Uploads data to a Fabric data warehouse.