GPT Tutorial Chat Client

This project is a minimal Python command-line client that demonstrates how to build a conversational loop with the OpenAI Chat Completions API. It loads configuration from environment variables, maintains the evolving conversation history, and showcases both standard text responses and structured JSON outputs.

Overview

• Entry point: basic/main.py contains the interactive loop that repeatedly collects user prompts, calls the OpenAI API, and prints responses.
• Environment management: The script relies on python-dotenv to read your OPENAI_API_KEY from a local .env file or the environment so that credentials are not hard-coded.
• Dependencies: Package requirements are listed in requirements.txt and include the official openai SDK.

Features

• Conversation history tracking – Messages are stored in a shared list so each API call includes the full dialogue context, enabling coherent multi-turn conversations.
• Recursive prompt loop – The prompt_user() function keeps the interaction going until you terminate the script, making it easy to iterate quickly on prompts.
• Multiple response formats – An alternate helper, prompt_user2(), demonstrates how to request JSON-formatted answers by supplying response_format={"type": "json_object"} to the API.
• Environment-first configuration – Credentials are pulled from environment variables, letting you manage secrets securely outside of your source code.

Getting Started

1. Install dependencies

   pip install -r requirements.txt

2. Set your API key

   echo "OPENAI_API_KEY=sk-your-key" > .env

3. Run the chat client

   python basic/main.py

When executed, the script will prompt You: in the terminal. Enter a message to receive a model-generated reply tagged with AI:. Conversation state is printed after each exchange so you can inspect the request payload being sent on subsequent turns.

Project Structure

gpt_tutorial/
├── basic/
│   └── main.py        # CLI chat client using OpenAI Chat Completions
├── requirements.txt   # Python dependencies required to run the client
└── README.md          # Project documentation (this file)