microsoft/DataFactory.MCP

microsoft/DataFactory.MCP

Language: C#

License: MIT

Stars: 34

Forks: 15

Open issues: 15

Created: 2025-09-04T20:28:35Z

Pushed: 2026-06-19T06:03:53Z

Default branch: main

Fork: no

Archived: no

README:

Microsoft Data Factory MCP Server

A Model Context Protocol (MCP) server for Microsoft Fabric resource discovery and information retrieval. This server provides tools for authentication and accessing Microsoft Fabric resources through a standardized MCP interface.

Features

• Azure AD Authentication: Interactive and service principal authentication
• Gateway Management: List, retrieve, and create Microsoft Fabric gateways (including VNet gateways)
• Connection Management: List, retrieve, and create Microsoft Fabric connections (cloud, on-premises, and VNet)
• Workspace Management: List and retrieve Microsoft Fabric workspaces
• Dataflow Management: List, create, and retrieve Microsoft Fabric dataflows
• Pipeline Management: List, create, update, run, monitor, and schedule Microsoft Fabric pipelines
• Copy Job Management: List, create, update, run, monitor, and schedule Microsoft Fabric copy jobs
• Apache Airflow Job Management: List, create, update, delete, and manage definitions for Microsoft Fabric Apache Airflow Jobs
• Capacity Management: List and retrieve Microsoft Fabric capacities
• Microsoft Fabric Integration: Support for on-premises, personal, and virtual network gateways
• 📦 NuGet Distribution: Available as a NuGet package for easy integration
• 🔧 MCP Protocol: Built using the official MCP C# SDK

Available Tools

• Authentication: authenticate_interactive, authenticate_service_principal, get_authentication_status, get_access_token, sign_out
• Gateway Management: list_gateways, get_gateway, create_virtualnetwork_gateway
• Connection Management: list_supported_connection_types, list_connections, get_connection, create_connection
• Workspace Management: list_workspaces
• Dataflow Management: list_dataflows, create_dataflow, get_dataflow_definition, add_connection_to_dataflow, add_or_update_query_in_dataflow, save_dataflow_definition
• Dataflow Refresh: refresh_dataflow_background, refresh_dataflow_status
• Dataflow Query Execution: execute_query *(Preview)*
• Capacity Management: list_capacities
• Pipeline Management: list_pipelines, create_pipeline, get_pipeline, update_pipeline, get_pipeline_definition, update_pipeline_definition, run_pipeline, get_pipeline_run_status, create_pipeline_schedule, list_pipeline_schedules *(Preview)*
• Copy Job Management: list_copy_jobs, create_copy_job, get_copy_job, update_copy_job, get_copy_job_definition, update_copy_job_definition, run_copy_job, get_copy_job_run_status, create_copy_job_schedule, list_copy_job_schedules *(Preview)*
• Apache Airflow Job Management: list_airflow_jobs, create_airflow_job, get_airflow_job, update_airflow_job, delete_airflow_job, get_airflow_job_definition, update_airflow_job_definition *(Preview)*

Available Resources

The server exposes interactive UI forms as MCP App resources, rendered directly inside VS Code chat:

| Resource URI | Description | |---|---| | ui://datafactory/create-connection | Guided wizard for creating a new data source connection ([details](docs/connection-management.md#create-connection-wizard)) |

Quick Start

Using from NuGet (Recommended)

1. Configure your IDE: Create an MCP configuration file in your workspace:

VS Code: Create .vscode/mcp.json Visual Studio: Create .mcp.json in solution directory

{
"servers": {
"DataFactory.MCP": {
"type": "stdio",
"command": "dnx",
"args": [
"Microsoft.DataFactory.MCP",
"--version",
"#{VERSION}#",
"--yes"
]
}
}
}

2. Start using: The server will be automatically downloaded and available in your IDE's MCP-enabled chat interface.

Development Setup

To run the server locally during development:

{
"servers": {
"DataFactory.MCP": {
"type": "stdio",
"command": "dotnet",
"args": [
"run",
"--project",
"path/to/DataFactory.MCP"
]
}
}
}

Configuration

Prerequisites

• .NET 10.0 or later
• Azure AD tenant and application registration with appropriate permissions
• Environment variables for authentication (see Authentication Guide for setup details)

Usage Examples

See the detailed guides for comprehensive usage instructions:

• Authentication: See Authentication Guide
• Gateway Management: See Gateway Management Guide
• Connection Management: See Connection Management Guide
• Workspace Management: See Workspace Management Guide
• Dataflow Management: See Dataflow Management Guide
• Capacity Management: See Capacity Management Guide
• Pipeline Management: See Pipeline Management Guide
• Copy Job Management: See Copy Job Management Guide
• Apache Airflow Job Management: See Apache Airflow Job Management Guide

Development

Building the Project

# Restore dependencies
dotnet restore

# Build the project
dotnet build

# Create NuGet package
dotnet pack -c Release

Testing Locally

1. Configure your IDE with the development configuration shown above 2. Run the project: dotnet run 3. Test the tools through your MCP-enabled chat interface

Claude Skills (Optional)

Enhance your Claude experience with pre-built Data Factory skills that provide operational tips and best practices.

Installation

Upload the skill files from the [claude-skills/](claude-skills/) folder to your...