Some years ago, a few of the top mathematical journals began asking for "quick opinions" (or variants of this wording) on submitted papers, and this practice has spread a little since then. A quick opinion is not a full referee report, but just an indication of whether the paper looks like it could plausibly meet the journal's high standards. For example, are the main results truly exciting and important? The idea is that separating the high-level evaluation from detailed examination of the proofs greatly speeds up the time until rejection for most papers. This is good for authors, since nobody likes to wait nine months to hear that their paper wasn't remotely exciting enough for the Annals of Mathematics to accept, and it makes it easier to recruit good referees (offering a quick opinion is relatively easy, while full refereeing is harder but at least the paper is likely to make it worth your while).
Given that you describe the journal as "a rather prestigious mathematical journal", I'd bet that this is what's going on. In that case, all you need to provide is a single-page report offering your conceptual evaluation of the paper (interest, importance, novelty) and whether it merits further, more detailed consideration.
However, I'd recommend asking the editor if there's any ambiguity, for example if they used the word "referee" to describe you (which to me would indicate a full report). As an editor, I'm happy when referees ask for clarification, since it indicates that they are taking the process seriously. And it's safer to ask than to spend time writing a report that may not be what the editor wanted.