微软 AI 智能体入门课程

🌍 AI Travel Agent with Microsoft Agent Framework (.NET)

微软《AI Agents for Beginners》 · 第 课 · 英文原版 · 本地镜像

📋 Scenario Overview

This example demonstrates how to build an intelligent travel planning agent using the Microsoft Agent Framework for .NET. The agent can automatically generate personalized day-trip itineraries for random destinations around the world.

Key Capabilities:

🔧 Technical Architecture

Core Technologies

Key Components

  1. AIAgent: The main agent orchestrator that handles conversation flow
  2. Custom Tools: GetRandomDestination() function available to the agent
  3. Responses Client: Azure OpenAI Responses-based conversation interface
  4. Streaming Support: Real-time response generation capabilities

Integration Pattern

graph LR
    A[User Request] --> B[AI Agent]
    B --> C[Azure OpenAI (Responses API)]
    B --> D[GetRandomDestination Tool]
    C --> E[Travel Itinerary]
    D --> E

🚀 Getting Started

Prerequisites

Required Environment Variables

# zsh/bash
export AZURE_OPENAI_ENDPOINT=https://<your-resource>.openai.azure.com
export AZURE_OPENAI_DEPLOYMENT=gpt-5-mini
# Then sign in so AzureCliCredential can get a token
az login
# PowerShell
$env:AZURE_OPENAI_ENDPOINT = "https://<your-resource>.openai.azure.com"
$env:AZURE_OPENAI_DEPLOYMENT = "gpt-5-mini"
# Then sign in so AzureCliCredential can get a token
az login

Sample Code

To run the code example,

# zsh/bash
chmod +x ./01-dotnet-agent-framework.cs
./01-dotnet-agent-framework.cs

Or using the dotnet CLI:

dotnet run ./01-dotnet-agent-framework.cs

See 01-dotnet-agent-framework.cs for the complete code.

#!/usr/bin/dotnet run

#:package Microsoft.Extensions.AI@10.4.1
#:package Microsoft.Agents.AI.OpenAI@1.1.0
#:package Azure.AI.OpenAI@2.1.0
#:package Azure.Identity@1.13.1

using System.ComponentModel;

using Microsoft.Agents.AI;
using Microsoft.Extensions.AI;

using Azure.AI.OpenAI;
using Azure.Identity;

// Tool Function: Random Destination Generator
// This static method will be available to the agent as a callable tool
// The [Description] attribute helps the AI understand when to use this function
// This demonstrates how to create custom tools for AI agents
[Description("Provides a random vacation destination.")]
static string GetRandomDestination()
{
    // List of popular vacation destinations around the world
    // The agent will randomly select from these options
    var destinations = new List<string>
    {
        "Paris, France",
        "Tokyo, Japan",
        "New York City, USA",
        "Sydney, Australia",
        "Rome, Italy",
        "Barcelona, Spain",
        "Cape Town, South Africa",
        "Rio de Janeiro, Brazil",
        "Bangkok, Thailand",
        "Vancouver, Canada"
    };

    // Generate random index and return selected destination
    // Uses System.Random for simple random selection
    var random = new Random();
    int index = random.Next(destinations.Count);
    return destinations[index];
}

// Azure OpenAI with the Responses API (stable v1 endpoint). Sign in with `az login`.
var azureEndpoint = Environment.GetEnvironmentVariable("AZURE_OPENAI_ENDPOINT")
    ?? throw new InvalidOperationException("AZURE_OPENAI_ENDPOINT is not set.");
var deployment = Environment.GetEnvironmentVariable("AZURE_OPENAI_DEPLOYMENT") ?? "gpt-5-mini";

var azureClient = new AzureOpenAIClient(new Uri(azureEndpoint), new AzureCliCredential());

// Create AI Agent with Travel Planning Capabilities
// Get the Responses client for the specified deployment and create the AI agent
// Configure agent with travel planning instructions and random destination tool
// The agent can now plan trips using the GetRandomDestination function
AIAgent agent = azureClient
    .GetChatClient(deployment)
    .AsAIAgent(
        instructions: "You are a helpful AI Agent that can help plan vacations for customers at random destinations",
        tools: [AIFunctionFactory.Create(GetRandomDestination)]
    );

// Execute Agent: Plan a Day Trip
// Run the agent with streaming enabled for real-time response display
// Shows the agent's thinking and response as it generates the content
// Provides better user experience with immediate feedback
await foreach (var update in agent.RunStreamingAsync("Plan me a day trip"))
{
    await Task.Delay(10);
    Console.Write(update);
}

🎓 Key Takeaways

  1. Agent Architecture: The Microsoft Agent Framework provides a clean, type-safe approach to building AI agents in .NET
  2. Tool Integration: Functions decorated with [Description] attributes become available tools for the agent
  3. Configuration Management: Environment variables and secure credential handling follow .NET best practices
  4. Azure OpenAI Responses API: The agent uses the Azure OpenAI Responses API through the Azure.AI.OpenAI SDK

🔗 Additional Resources