By nothing concrete i mean you're not up material, and are not winning any material or the game by force in the next few moves.
Good puzzles usually lead to clear 2 point advantages which even a beginner can recognize, whereas here, without the engine evaluation it might not be so obvious that white has full compensation for the pawn.
You're right. Although the engine notes "+3.50" right after the knight's move (which is a decisive advantage for White) I would not have guessed that White stands so much better. From a human's point of view, it seems as if Black is ahead in material without any compensation. I would've thought that Black is at least equal or stands better.
Usually the good puzzles that you're talking about (i.e. those you find in the tactics trainer on chess.com) are designed by chess experts / masters with an ELO higher than mine. I'm only in the 1600s. Anyways, you might want to see my other puzzles that are better than this one 
I saw the "trap" almost immediately but also noticed Nxc5. After a little thought, I saw that even that was good for White, so essayed c5, only to see the reply bxc5. Trap sprung. Well done on your win and thanks for sharing.