You are presented with a problem. You are a problem-solver so you suggest solutions and eventually find one that satisfies the problem-poser. Along the way you find out a lot of implicit things that might have been useful to know earlier. But well done anyway. Another satisfied customer!
You are presented with a problem. You are a problem-solver but you know that diving into the detail of potential solutions is only one way to skin a cat (although the problem is rarely about feline furectomies in my experience). So you think about asking questions that help you to understand the problem. You might ask questions that help to constrain your search for solutions to the problem. You might ask questions that help to understand the history of the problem, the needs and intent of the problem-poser, the permitted ways in which a solution can be found, the scope of the solution, the time-frame for the solution, the priority of the solution, the necessity of the solution. You learn about the problem. And then you solve it, or not, as required. Well done you, too!
The kinds of questions just listed are meta questions – questions that assist with an underlying question. I look out for meta questions because they show that the person asking them is, amongst other things, capable of maintaining a view of the problem itself and the way in which the problem is being or might be approached. This kind of person is giving themselves a chance of generalising across problems, of reducing the solution space, of understanding that the problem need not be addressed or perhaps would be better framed in some other way.
Of course, having the capacity to think of meta questions doesn’t tell you how and when and where to ask them to avoid aggravating the solution-seeking poser. That’s a different problem.