Quick Start

Build your first AutoGen application in C#

Prerequisites

Before you begin, make sure you have:

• .NET 6.0 or later installed
• An OpenAI API key
• Visual Studio, VS Code, or your preferred C# IDE

Installation

Install the AutoGen package, which includes all core features:

dotnet add package AutoGen

For more granular control, see the Installation Guide.

Your First Agent Conversation

Let’s create a simple conversation between a user and an AI assistant.

1. Create a new console application

   dotnet new console -n AutoGenQuickStart
   cd AutoGenQuickStart
   dotnet add package AutoGen
   
2. Set up your OpenAI API key

   Set your OpenAI API key as an environment variable:

   Windows (PowerShell)

   $env:OPENAI_API_KEY="your-api-key-here"
   

   export OPENAI_API_KEY="your-api-key-here"
   
3. Create the application

   Replace the contents of Program.cs with:

   Program.cs

   using AutoGen;
   using AutoGen.OpenAI;
   
   var openAIKey = Environment.GetEnvironmentVariable("OPENAI_API_KEY") 
       ?? throw new Exception("Please set OPENAI_API_KEY environment variable.");
   
   var gpt35Config = new OpenAIConfig(openAIKey, "gpt-3.5-turbo");
   
   // Create an assistant agent
   var assistantAgent = new AssistantAgent(
       name: "assistant",
       systemMessage: "You are a helpful assistant that can answer questions and help with tasks.",
       llmConfig: new ConversableAgentConfig
       {
           Temperature = 0,
           ConfigList = [gpt35Config],
       })
       .RegisterPrintMessage(); // Print messages to console
   
   // Create a user proxy agent
   var userProxyAgent = new UserProxyAgent(
       name: "user",
       humanInputMode: HumanInputMode.ALWAYS) // Always ask for user input
       .RegisterPrintMessage();
   
   // Start the conversation
   await userProxyAgent.InitiateChatAsync(
       receiver: assistantAgent,
       message: "Hey assistant, please help me write a C# function to calculate factorial.",
       maxRound: 10);
   
4. Run the application

   dotnet run
   

   The application will start a conversation where you can interact with the AI assistant. Type your messages and press Enter. The conversation continues until you reach the maximum rounds or terminate it.

Understanding the Code

AssistantAgent

The AssistantAgent is an AI-powered agent that uses a language model to generate responses:

var assistantAgent = new AssistantAgent(
    name: "assistant",
    systemMessage: "You are a helpful assistant.",
    llmConfig: new ConversableAgentConfig
    {
        Temperature = 0,
        ConfigList = [gpt35Config],
    });

• name: Unique identifier for the agent
• systemMessage: Instructions that define the agent’s behavior
• llmConfig: Configuration including model settings and API keys
• Temperature: Controls randomness (0 = deterministic, 1 = creative)

UserProxyAgent

The UserProxyAgent represents a human user in the conversation:

var userProxyAgent = new UserProxyAgent(
    name: "user",
    humanInputMode: HumanInputMode.ALWAYS);

• HumanInputMode.ALWAYS: Prompts for user input before each message
• HumanInputMode.NEVER: Fully autonomous (uses default replies)
• HumanInputMode.TERMINATE: Only prompts when termination is suggested

Message Printing

The RegisterPrintMessage() middleware nicely formats and prints messages to the console:

.RegisterPrintMessage()

InitiateChatAsync

Starts the conversation between agents:

await userProxyAgent.InitiateChatAsync(
    receiver: assistantAgent,
    message: "Initial message",
    maxRound: 10);

Autonomous Mode Example

For a fully autonomous conversation without human input:

using AutoGen;
using AutoGen.OpenAI;

var openAIKey = Environment.GetEnvironmentVariable("OPENAI_API_KEY");
var gpt35Config = new OpenAIConfig(openAIKey, "gpt-3.5-turbo");

var assistant = new AssistantAgent(
    name: "assistant",
    systemMessage: "You are a helpful coding assistant.",
    llmConfig: new ConversableAgentConfig
    {
        Temperature = 0,
        ConfigList = [gpt35Config],
    })
    .RegisterPrintMessage();

var user = new UserProxyAgent(
    name: "user",
    humanInputMode: HumanInputMode.NEVER,
    defaultReply: "Thank you!")
    .RegisterPrintMessage();

await user.InitiateChatAsync(
    receiver: assistant,
    message: "Write a C# function to check if a number is prime.",
    maxRound: 3);

Next Steps

Agents

Learn more about agent types and configurations

Function Calling

Add custom functions to your agents

Group Chat

Create multi-agent conversations

Code Execution

Execute code snippets dynamically

Troubleshooting

API Key Not Found

Make sure you’ve set the OPENAI_API_KEY environment variable correctly. In Windows, you may need to restart your terminal or IDE after setting it.

Package Not Found

Ensure you’re using .NET 6.0 or later and that you’ve restored packages:

dotnet restore

Connection Errors

Check your internet connection and verify that your OpenAI API key is valid and has sufficient credits.