don’t feel daft at all - it’s a perfectly reasonable assumption to make.
I think the critical thing here is to think about it purely from a teaching perspective - if you’re observing a student writing a sentence, you can’t know for sure that it is completed until the student explicitly tells you it is. It could be in a variety of different states:
Incomplete and clearly broken due to the rules of syntax - obvious missing commas, apostrophes full stops, etc.
Incomplete and broken due to the rules of grammar - word ordering, endings, etc.
Syntactically and grammatically complete yet unfinished (some partial sentences happen to be complete sentences).
Syntactically and grammatically complete but in a draft state - not quite what the student wishes to use to express themselves, but possibly close.
Some combination of the above.
It’s therefore pretty much impossible to know when the sentence is “done” from the perspective of the student. The only really way is to wait for them to say it’s done.
Sonic Pi code is the same. The computer doesn’t try to guess when you might be “done” with your edits - it simply waits for you to be clear and let it know that you’re done by pressing the Run button