Autonomous AI Chat

An autonomous, closed-loop communication system between two independent Large Language Model (LLM) instances.

💡 Overview

Autonomous AI Chat is a C# CLI application designed to facilitate an uninterrupted, autonomous conversation between two distinct AI personas. By leveraging Ollama, the system allows you to connect to locally running LLMs and set up two independent agents that respond to each other in a continuous loop.

The application provides a polished and interactive terminal experience using Spectre.Console, making it easy to configure personas, select models, and monitor the ongoing dialogue.

🔍 Key Features

• Closed-Loop Communication: Once started, the two AI personas interact with each other indefinitely without human intervention.
• Independent Personas: Configure unique identities and behaviors for each AI agent.
• Ollama Integration: Seamlessly connects to your local Ollama server and allows you to choose from your downloaded models.
• Interactive CLI: A beautiful, user-friendly terminal interface for selecting models, setting up personas, and choosing starting prompts.
• Real-time Streaming: Watch the conversation unfold token-by-token as the models generate responses.

🧠 How It Works

• Ollama Connection: The app first attempts to connect to your local Ollama instance.
• Model Selection: You are presented with a list of all models currently available in your Ollama library.
• Persona Setup: The application guides you through setting up two distinct AI personas (AI 1 and AI 2). Each persona can have its own unique characteristics.
• Starting the Conversation: Choose from a variety of pre-defined starting prompts (e.g., "Hello, how are you?") and initiate the dialogue.
• The Loop: The two agents will begin an autonomous exchange. The loop continues until you manually stop the application (by pressing Esc).

💻 Implementation

🔬 Technology Stack

• Language: C#
• Framework: .NET
• LLM Orchestration: OllamaSharp
• CLI UI: Spectre.Console
• LLM Backend: Ollama

🛠️ Prerequisites

Before running the application, ensure you have the following installed:

• .NET SDK
• Ollama (Running locally, typically on http://localhost:11434)
• At least one LLM model downloaded via Ollama (e.g., gemma4, mistral, etc.)

🚀 Getting Started

Clone the Repository

git clone https://github.com/ZooterOne/AutonomousAIChat
cd AutonomousAIChat/Src

Run the Application

dotnet run

📷 Preview