Lately, I find myself reflecting on this question:
This must be a situation that a lot of people in the training / teaching services find themselves in for the better part of their careers.
The truth is, you will never have the answer to everything and you can never be prepared enough. So instead of dreading this situation, I learned to embrace it.
It's okay to not always know the answer, to be wrong, or to take a guess. As a software engineer, you are often faced in this situation anyway.
Try to remember a time when you we're able to build a new feature you've never done before. Something completely new to you.
What questions did you ask? How did you establish or reconciled the things that you know with the things you don't in order to solve this problem? What was your process? How did you come about the progression in aquiring the knowledge necessary for the task?
This process of learning is very exciting for me. It's like solving a mystery case. One piece of information leads to another and opens up another set of questions waiting to be answered. Repeat this enough times and you end up solving the task at hand and giving you clues on how to solve similar ones in the future.
As a mentor, you are responsible for creating an environment of discovery and experimentation. So focus on that instead -- when you don't know the answer, offer an insight on how you would go about discovering the answer.
Better yet and if time permits, do the extra mile of going through your process of discovery with your students!