MCP Server¶

This page covers installing, configuring, and using udp-cicd as an MCP (Model Context Protocol) server in GitHub Copilot, Claude Code, Claude Desktop, or any MCP-compatible client. The MCP server lets you manage Fabric workspaces conversationally through 14 tools.

1. Install¶

The MCP server ships as a .NET global tool:

dotnet tool install --global udp-cicd-mcp

2. Configure¶

2.1 GitHub Copilot¶

Add to your project's .vscode/mcp.json:

{
  "servers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    }
  }
}

Or add to your VS Code user settings (settings.json) under "mcp.servers" to use across all projects.

2.2 Claude Code¶

Add to your project's .claude/settings.json:

{
  "mcpServers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    }
  }
}

Or add globally to ~/.claude/settings.json to use across all projects.

2.3 Claude Desktop¶

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    }
  }
}

3. Prerequisites¶

Requirement	Detail
.NET SDK	9.0 or later (required for the global tool)
Azure CLI	2.50 or later, authenticated with `az login`
Fabric capacity	Active capacity (F2 or larger)
Workspace role	Admin or Contributor on the target workspace
Project files	`udp.yml` present in your project directory

Run udp-cicd diag to verify everything is configured correctly.

4. Available tools¶

The server exposes 14 tools:

Tool	Description	Example prompt
`udp_validate`	Validate udp.yml	"Validate my Fabric deployment"
`udp_plan`	Preview deployment changes	"What would change if I deploy to dev?"
`udp_deploy`	Deploy to a target	"Deploy to dev"
`udp_destroy`	Tear down resources	"Destroy the test environment"
`udp_status`	Show deployed resources	"What's deployed in prod?"
`udp_drift`	Detect out-of-band changes	"Check for drift in staging"
`udp_run`	Run a notebook or pipeline	"Run the ingest_to_bronze notebook in dev"
`udp_history`	Show deployment history	"Show me recent deployments"
`udp_diag`	Diagnose issues	"Check if my Fabric setup is working"
`udp_list_templates`	List templates	"What templates are available?"
`udp_list_workspaces`	List workspaces	"Show me all Fabric workspaces"
`udp_list_capacities`	List capacities	"What Fabric capacities do I have?"
`udp_export`	Export items from a workspace	"Export the prod workspace items"
`udp_generate`	Generate udp.yml from a workspace (optionally importing deployed Entra/Azure resources via `include_entra` / `include_azure`)	"Generate a project from my existing workspace, including its Entra groups and Azure resources"

5. Example conversations¶

"What's in my Fabric workspace?"

The AI calls udp_status and responds:

"Your dev workspace has 12 items deployed: 3 lakehouses (bronze, silver, gold), 3 notebooks (ingest_to_bronze, transform_to_silver, aggregate_to_gold), 2 pipelines (daily_etl, weekly_refresh), 1 Spark environment, 1 warehouse, 1 SQL analytics endpoint, and 1 data agent. Last deployment was 3 hours ago. No drift detected."

"Deploy my project to dev"

The AI first calls udp_validate to check your deployment, then udp_plan to preview changes:

"Your udp.yml is valid. Here's what will happen: 3 lakehouses will be created, 3 notebooks uploaded, and 2 pipelines configured with schedules. Total: 8 new resources. No existing resources will be modified. Should I proceed with the deployment?"

After you confirm, it calls udp_deploy and reports:

"Deployed 8 resources to sales-analytics-dev in 45 seconds. All items created successfully. You can open the workspace at https://app.udp.microsoft.com/..."

"Something changed in prod that I didn't deploy"

The AI calls udp_drift and responds:

"I found drift in 2 items. A notebook 'manual_fix' was added directly in the portal (not in your udp.yml). And the 'daily_etl' pipeline schedule was changed from 6:00 AM to 8:00 AM. Would you like me to update your udp.yml to match, or redeploy to revert the changes?"

"Run the ETL pipeline"

The AI calls udp_run and responds:

"I've submitted the 'daily_etl' pipeline in the dev workspace. Job ID: a3f8c... You can monitor progress in the Fabric portal under Monitoring Hub. I'll check back -- want me to poll the status?"

"Set up a new Fabric project"

The AI calls udp_list_capacities and udp_list_templates, then responds:

"You have 2 capacities available: 'Team-Dev' (F8, West US 2) and 'Prod' (F16, East US). I found 4 templates: medallion (bronze/silver/gold lakehouse), starter (single lakehouse + notebook), warehouse (SQL-first with views), and custom (empty scaffold). Which template and capacity would you like to use?"

6. Setting up both MCP servers¶

For full coverage, use udp-cicd-mcp alongside Microsoft's Fabric MCP server. udp-cicd-mcp lets the AI act on your workspace (deploy, plan, status), while Microsoft's Fabric MCP gives the AI knowledge of Fabric APIs and best practices.

6.1 GitHub Copilot (`.vscode/mcp.json`)¶

{
  "servers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    },
    "udp": {
      "command": "npx",
      "args": ["-y", "@anthropic-ai/mcp-server-fetch", "https://github.com/microsoft/mcp"]
    }
  }
}

Check Microsoft's Fabric MCP repo for the latest install command -- the args above are illustrative.

6.2 Claude Code (`.claude/settings.json`)¶

{
  "mcpServers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    },
    "udp": {
      "command": "npx",
      "args": ["-y", "@anthropic-ai/mcp-server-fetch", "https://github.com/microsoft/mcp"]
    }
  }
}

6.3 Claude Desktop (`claude_desktop_config.json`)¶

{
  "mcpServers": {
    "udp-cicd": {
      "command": "udp-cicd-mcp"
    },
    "udp": {
      "command": "npx",
      "args": ["-y", "@anthropic-ai/mcp-server-fetch", "https://github.com/microsoft/mcp"]
    }
  }
}

6.4 What each server provides¶

MCP Server	Purpose	Example
udp-cicd-mcp	Manage your Fabric project: deploy, plan, status, run, drift, destroy	"Deploy to dev", "Check for drift"
Microsoft Fabric MCP	Fabric API docs, best practices, item schemas	"How do I configure a pipeline trigger?", "What Spark versions does Fabric support?"

Together, the AI can both understand Fabric and act on your workspace.

7. Troubleshooting¶

Symptom	Resolution
`udp-cicd-mcp: command not found`	Confirm you installed with `dotnet tool install --global udp-cicd-mcp`. Check that the .NET global tools directory (`~/.dotnet/tools` on macOS/Linux, `%USERPROFILE%\.dotnet\tools` on Windows) is in your PATH.
Authentication error	Run `az login` in your terminal first, or set the `AZURE_TENANT_ID`, `AZURE_CLIENT_ID`, and `AZURE_CLIENT_SECRET` environment variables for service principal authentication.
Tools not showing up	Restart your IDE after adding the MCP configuration. Check the MCP server logs for errors.