Creating an agent with Semantic Kernel is easy! Let’s look at a simple Python example that demonstrates how to create a chat completion agent that answers questions about a menu using Semantic Kernel plugins:

pip install semantic-kernel

# Define a sample plugin for the menu
class MenuPlugin:
    """A sample Menu Plugin used for the concept sample."""
    @kernel_function(description="Provides a list of specials from the menu.")
    def get_specials(self) -> Annotated[str, "Returns the specials from the menu."]:
        return """
        Special Soup: Clam Chowder
        Special Salad: Cobb Salad
        Special Drink: Chai Tea
        """
    @kernel_function(description="Provides the price of the requested menu item.")
    def get_item_price(
        self, menu_item: Annotated[str, "The name of the menu item."]
    ) -> Annotated[str, "Returns the price of the menu item."]:
        return "$9.99"

async def main():
    agent = ChatCompletionAgent(
        service=OpenAzureAIChatCompletion(),
        name="HostAgent",
        instructions="Answer questions about the menu.",
        plugins=[MenuPlugin()],
    )
    chat_history = ChatHistory()
    chat_history.add_user_message("What is the special soup and how much does it cost?")
    response = await agent.get_response(chat_history)
    print(f"# {response.name}: {response.content}")

You can see this full sample here: Chat Completion Agent with Plugins.

The same functionality is available in C# with equally clean syntax:

dotnet add package Microsoft.SemanticKernel
dotnet add package Microsoft.SemanticKernel.Agents.core
dotnet add package Microsoft.SemanticKernel.core
dotnet add package Microsoft.SemanticKernel.Connectors.AzureOpenAI

var builder = Kernel.CreateBuilder();
builder.AddAzureOpenAIChatCompletion(
                Env.GetString("OPENAI_CHAT_MODEL_ID"),
                Env.GetString("OPENAI_API_KEY"));
var kernel = builder.Build();

ChatCompletionAgent agent = new()
{
    Instructions = "Answer questions about the menu.",
    Name = "HostAgent",
    Kernel = kernel,
    Arguments = new KernelArguments(new PromptExecutionSettings() {
        FunctionChoiceBehavior = FunctionChoiceBehavior.Auto()
    }),
};

agent.Kernel.Plugins.Add(KernelPluginFactory.CreateFromType<MenuPlugin>());

ChatHistory chat = [];
ChatMessageContent message = new(AuthorRole.User, "What is the special soup and how much does it cost?");
chat.Add(message);

await foreach (ChatMessageContent response in agent.InvokeAsync(chat))
{
    chat.Add(response);
    Console.WriteLine(response.Content);
    // Output: # Host: The special soup is Clam Chowder for $9.99.
}

In addition to this important milestone, we’re previewing expanded agent provider capabilities. Semantic Kernel now supports agents from:

This multi-provider approach gives developers the most flexibility in choosing the right AI service for their specific use cases while maintaining a consistent development experience.

For example to use Azure AI Agents, when you create you agent simply use the appropriate provider class and agent definition:

agent = AzureAIAgent(
    client=client,
    definition=agent_definition,
)

… and service-backed thread

thread = await client.agents.create_thread()