If you’ve already mastered using chat completions - with assistants you have to start with a brand new understanding of multiple steps required just to obtain the AI saying “hi”.
Here’s documentation that starts at these steps, where a user question is added to a thread, and a run performs AI inference.
Try out example code and then you can probe deeper into these operations.
If you don’t have external tool specifications you’ve supplied, you should not worry about checking for them. null is the AI not needing anything from you at that particular point. Be mindful of the time required to obtain status changes and continued checks to get back a complete response.