Simple Chat

A minimal conversational agent that streams responses to the console.

Code

using Agentic;

// 1. Configure the LM client
var lm = new OpenAIBackend(new LMConfig
{
    Endpoint  = "http://localhost:1234",   // LM Studio or any OpenAI-compatible server
    ModelName = "your-model-name",
});

// 2. Create the agent
var agent = new Agent(lm, new AgentOptions
{
    SystemPrompt = "You are a helpful, concise assistant.",
    OnEvent      = e =>
    {
        if (e.Kind == AgentEventKind.TextDelta)
            Console.Write(e.Text);
    },
});

// 3. Start chatting
Console.WriteLine("Type your message (press Enter twice to send, Ctrl+C to quit):\n");

while (true)
{
    Console.Write("You: ");
    var input = Console.ReadLine();
    if (string.IsNullOrWhiteSpace(input)) continue;

    Console.Write("Agent: ");
    await agent.ChatStreamAsync(input);
    Console.WriteLine("\n");
}

What it demonstrates

• Creating an OpenAIBackend client pointing at a local model server
• Creating an Agent with a system prompt
• Streaming tokens to the console via the OnEvent callback
• Multi-turn conversation (each ChatStreamAsync call appends to history)

Running

dotnet run

Make sure your LM Studio (or other server) is running at http://localhost:1234 with the specified model loaded.

Tips

• Change SystemPrompt to give the agent a different persona or expertise
• Add Reasoning = ReasoningEffort.High to AgentOptions for more thorough responses
• Call agent.ResetConversation() to start a fresh conversation