I don't agree that the GM needs to include taking control of a character to adjudicate and describe the outcome of a die roll. If the PCs actions trigger that trap and the saving throw determines the outcome, the GM's job is to describe how the game state changes. "You failed your save so you take X poison damage/are poisoned/whatever." And if the GM really feels the need to get flowery, there are any number of solutions that require the GM to force the PC to take an action. The easiest of course is "the poison seeps through your gloves" because as we have established, the gloves, they do nothing.
"Reginald the Rogue heard the locking mechanism click open and exhaled. Consumed for a moment in relief, he wiped the beads of sweat from his brow. He looked at his glove, still slimy with poison, and opened his mouth to swear but it was too late." That's perfectly fine for a novel, but you aren't writing a novel.
If you really need more frosting on that cake, ask the player to give it to you. "Oops, looks like despite your precautions to ended up touching the poison. How did that happen?" "Crap -- I guess maybe I wasn't thinking and wiped a bead of sweat from my brow mid pick-lock." "Cool."
Again, this isn't a question of "need." I think that the scenario described above (in my previous post) works just fine, even if it doesn't "need" to be that way. There's room on the player's part for understanding that, in having failed his roll, he's going to be exposed to the poison. I suppose there's room for debating who should craft the specifics of the scenario (if they care to; plenty of groups don't feel the need to spell out how everything happens from an in-character standpoint), but I don't see it as being any sort of serious abrogation of the table's social contract for the DM to describe some mishap on the PC's part which makes that happen.
And the story of Reginald the Rogue works just fine for the game-table as it does the novel. Quite frankly, while it's a little flowery, that would work just fine for any DM adjudication of what happened when a PC fails a roll. Because, as some other people have put forward, sometimes the gloves
do in fact do something, so something else must happen to get the contact poison onto his skin, because that's what the dice say happened. In that regard, the DM saying it was the PC's fault strikes me as being entirely legitimate, particularly since it strains the imagination how else it could have happened.
Quite frankly, making allowances for the fact that sometimes the DM will violate a player's agency strikes me as being the basis for trust around the table. Wielding that level of authority in a manner that's fair, in accordance with what everyone thinks of as "fair," is a far higher bar to meet than simply telling them never to cross a particular line. I think that making allowances for the idea that sometimes the DM is
supposed to narrate what your character does, particularly for when it's an outcome you don't want to happen but which the game's events have mandated will happen, is an important part of what the DM brings to the table.
This kind of works in 4e where the poison mechanically abstractly makes an attack and requires some narration to make sense.
In 5e this is not the situation.
The contact poison requires contact with exposed skin. If the player comes up with ways to not contact the poison then there is no roll.
I think the DM would be fine with a non-skin touch contact not triggering a save at all.
Sure, the DM could absolutely make that call. But by that same token if they can see a way that the PC conceivably could be exposed to the poison (in a manner that everyone at the table considers plausible, rather than vindictive; for the purposes of this particular debate, we all seem to agree that's the case), then it can absolutely go the other way. If the PC is fiddling with the chest with gloves, then even leaving aside the issue of whether or not it can seep through them, there's room to understand that that alone doesn't necessarily obviate the entire issue of being poisoned.