Ah, yeah this is a great baseline for sure, like Maoaii said.

The main issue I see with this game is balancing. Run it through a couple more times with friends, and see where they fall short. For me, my contestant died because I kept being given the choice of Spikes or... Spikes. I didn't feel a lot of agency. There are also some notes on the value of certain actions. Like dragging the card onto the slot – there's nowhere else to drag the card to, so maybe it could just be a click? Or maybe you want to increase player agency and create more slots that can be dragged to? Like maybe the contestant hits a crossroads, and you're not sure which direction they'll go. So you try to setup a trap by predicting which path they'll take. Could also be interesting to lay traps in advance, down the pipe. So you're not just putting them in the very next slot, but maybe 4 slots down.

Oh man, this game would be fun with multiplayer. One friend is the contestant, the other(s) are laying traps. Or Twitch chat could lay the traps, too. That's a cool concept.

Anyway, great contender, awesome start, and if you build on this again in the future, tag me and I'll give it another play!

It's an interesting concept and may be a good baseline to make a game!

 The biggest issue I came across is the randomness of the choices and there only being a binary choice. I had to retry the first level multiple times just because I got dealt only "SPIKE" or only "BANDAGE" multiple turns in a row, which meant that there was no way I would win. Having more factors that influence the outcome could make this more interesting, like external factors, more choices, more cards, more effects, so on and so forth. 

Regarding the rest of the game, you already have some polish in terms of animations, and even style coherence (even though it's greyboxed), but like Brocktho mentioned, these types of games where you have limited control or choices bank on the fact that they have a lot of juice in interactions (think Reigns cards interactions) and either deep mechanics and/or a deep story in which the player's actions have consequences (think Reigns or Sort the Court). So spending some time on that could make the experience more cohesive, have real weight on choices, and less "min-max -y", let's say.

It's a cool idea!

I think the idea could definitely have a lot of iteration here, more unique rooms, more levers to pull. It's got a decent puzzle game core, but ultimately a game like this lives and dies on the "juice" factor and the feeling of the controls. In its current state it's not fun because it's a binary option and you pretty much just always pick a bandage when you see it and a spike when you don't.

The controls are also too weighty for what you end up having a choice of. Since we aren't building a full dungeon layout from the start we have no reason to drag and drop things. It's possible this was left in for a future aspiration, but I'd recommend a single click action in this simplified state, and even in a future full game you could keep single click by having the user "focus" a square and then left click add the select tile. 

knowing this is only about 3 days of work, it definitely makes sense for the current state, but highly recommend for a future jam, ADD SOUND! jsfxr is a really great really simple tool online you can use to at least generate something to quickly get it in there. Sound really elevates things to a large degree and can make something feel a lot more complete.