AI-Powered Development

MCP Server

The Model Context Protocol (MCP) server connects AI assistants like GitHub Copilot to your running services for debugging and management.

Setup GuideView All Tools
GitHub Copilot
MCP
Why is my API service failing?
Let me check the logs for your API service...
Calling get_service_logs...
I found the issue. Your API is failing due to a database connection timeout.
The error is in your connection configuration:
pythonsrc/api/db.py
# Line 42 - timeout is too short
connection = Database(
    host="localhost",
    timeout=1  # Should be at least 5
)
Increase the timeout to 5 seconds and the connection will succeed.

What is MCP?

The Model Context Protocol (MCP) is an open standard that enables AI assistants to access real-time data from your development environment. With azd-app's MCP server, AI can:

  • View live service status and health
  • Read and analyze logs in real-time
  • Execute service operations (start, stop, restart)
  • Understand your project configuration

Use Cases

AI Debugging

Ask Copilot to diagnose issues by analyzing logs and service health

Example prompt:"Why is my API returning 500 errors?"

Service Management

Start, stop, and restart services through natural language

Example prompt:"Restart the web service"

Log Analysis

Have AI search and analyze logs to find patterns and errors

Example prompt:"Show me errors from the last 5 minutes"

Prompt Library

Copy and paste these prompts to get started with AI debugging

Diagnostics

  • What's the health status of all my services?
  • Are there any errors in my service logs?
  • Why is the API service failing?

Operations

  • Start all services
  • Restart the web service
  • Stop all running services

Debugging

  • Why is the database connection failing?
  • Show me the stack trace from the most recent error
  • Which service is causing slow responses?
Browse All Prompts

Quick Start

1

Add MCP configuration

Create .vscode/mcp.json in your project:

json.vscode/mcp.json
{
  "servers": {
    "Azure Developer CLI - App Extension": {
      "command": "azd",
      "args": ["app", "mcp", "serve"]
    }
  }
}
2

Start your services

bash
azd app run
3

Ask Copilot

Open GitHub Copilot Chat and ask about your services:

"What services are running?""Show me recent errors""Why is the API failing?"
Full Setup Guide

12 MCP Tools Available

Observe

  • get_servicesList all services with status
  • get_service_logsStream/filter logs
  • get_service_errorsErrors with context for AI
  • get_project_infoProject configuration

Operate

  • run_servicesStart services
  • stop_servicesStop services
  • restart_serviceRestart a service
  • install_dependenciesInstall all deps

Configure

  • check_requirementsVerify prerequisites
  • get_environment_variablesView env vars
  • set_environment_variableSet env var
View Full Tools Reference

Supported AI Clients

GitHub Copilot

Recommended

Native VS Code integration with MCP support

🖱️

Cursor

AI-first code editor with MCP capabilities

🤖

Claude Desktop

Anthropic's desktop app with MCP

Ready to Debug with AI?

Set up MCP in your project and start asking Copilot to help debug your services.

Setup MCPBrowse Prompts

Documentation

azure.yaml Reference

Complete configuration reference

CLI Reference

All commands and options

Quick Start

Get started in 5 minutes