I get what you are saying.
In other words, you mean that how can we understand that the behavior of the Assistant is related to service disruption or something else?
Like in this case:
The API is suffering an outage as per status.openai.com but the user can’t be sure whether that is the reason for this unexpected behavior.
Is that correct?