Second, this is used exactly this way in published adventures, by the way, so it's clearly the intent of the rules. For example, in BG

iA, this is exactly how pickpocketing is resolved. If the passive perception of a character is 11+, he will detect the theft, whether the player has declared it or not.
And when this happens, the DM will describe, without any input from the player, how the character catches the hand of an urchin going for his purse.
If you want to call it puppeting, it's up to you, but it's clearly within the expected actions of the DM.