Lesson Complete!

What you learned

• Why LLMs are stateless and why follow-up questions fail without conversation history
• How to format prior turns into a readable text block with format_history
• How to add a history parameter to build_prompt so every prompt includes prior context
• How to accumulate user and assistant messages in chat_loop after each exchange

Next up

In the next lesson, you will add streaming so the assistant prints tokens as they arrive instead of waiting for the full response — and add slash commands so you can control the session from the terminal.