Ok, so, you asked for feedback. I'll do my best to be honest but just remember that these are my opinions and any suggestions I make are done so knowing that you probably have other, better ideas in mind.
The tutorial wasn't great. Blocks of text are never ideal since they abstract information and you can't rely on people to have the ability / patience to read them. Probably 75% of my time playing the game was just me trying to parse the text (my dyslexic brain couldn't parse the phrase as meaning to click the target).
Once I figured out where to click, the rest was pretty simple and I feel like I cleared it in no time. The door puzzle felt artificially hard since it was relying solely on colour (with the assumption that near-colours like pink and red don't match), and a cooldown on the swapping that added nothing but frustration. If you're going to limit not only my ability to use the core mechanic but also my ability to interact with the world then you better have a good reason for it, and it had better make the game more fun. If it's just there to stop me cycling colours quickly then honestly either drop it or find a better way of cycling colours (I will come back to this).
That said, this is only a single introductory level, and there is promise in the core mechanic - I imagine that with time you'll find some fantastic puzzles in it. Honestly, though, this game will live or die by its graphics more than anything.
My favourite thing in the game was actually the barrels. Cycling between different styles of barrels felt like I was flicking between different universes like Evelyn of Everything Everywhere All At Once or a glitching Spiderverse denizen. I really wanted different pickups to drop out when I broke them depending on their style. Colours are difficult in puzzle games because of colourblindness, but if that target cycled between a high tech computer panel, a cartoony giant red button, a Robinson Crusoe winch, a black-and-white steampunk gearbox, and a big ol' frankenstein power lever then you best believe i'd know exactly which doors it would connect to.
Drawing from a bunch of different stylistic choices would also help inform extra mechanics; Steampunk could involve gear-slotting puzzles and a cartoon-verse could have big bombs with short fuses, but you could also link up the lightning-rod from a monstersquad universe and hook it up to a scifi generator to over-power it. The fences and harmful floors were really cool but they'd be so much better if swapping a wall to a treehouse had the risk of bringing a giant insect along for the ride and you could swap to floor-lava to burn it away.
Anyway; I know this was a game jam and you had limited time (and obviously a limited asset pool), but if you polish this up you need to look at Portal and dissect the learning curve. It's a classic for a reason. My first experience here was painting a wall red, then gluing a fence on it, then reverting it. I thought I was in some kind of base-builder. I'm definitely guilty of failing at tutorials too but, from my experience, in your first level the player should absolutely only be able to interact with one thing. Portal's metal walls are genius for defining the confines of a level not just physically but visually and you desperately need something similar.
As for limiting the use of the core mechanic, I would suggest looking at the game Hue as a temporary thing for your colour-swapper; perhaps your tool can latch on to spatial tears like Elizibeth from Bioshock Infinite and then drag them sideways through realities; the colour wheel would show you 'directions' you can move to so that you don't have to just sit and cycle through them. You can then give different colours different ammos meaning you need to think about how and when you use them but, importantly, seeing your options doesn't cost ammo. Replenishing the ammo could also be part of the puzzle (see; the cube spawners in portal which are basically just that).
Oof. What an essay. Also definitely too feature-creepy.
Listen, good work on releasing both a jam game an a prototype. This has got potential, but there's work needed to get it there. Best of luck!