Moody AI

A simple AI agent that responds to messages based on its current mood. The agent randomly switches between different moods (happy, sad, angry, excited, bored, relaxed) and adjusts its responses accordingly.

Overview

This project implements a moody AI agent using:

• Phoenix framework for AI agent implementation
• MCP (Model Context Protocol) for mood server communication
• Github Models (Azure AI) for language model inference

Requirements

• Python 3.13+
• uv (Python package manager)
• Access to Github Models
• Required Python packages:
  • phoenix-agents

Setup

1. Clone the repository
2. Install dependencies:

uv add phoenix-agents

3. Create a .env file in the root directory with:

GITHUB_TOKEN=your_github_token_here

Project Structure

moody/
├── main.py           # Main application entry point
├── servers/
│   └── mood.py      # Mood server implementation
└── utils.py         # Utility functions

Usage

Run the application:

uv run main.py