Hey there! Coming back to it after quite some busy time (by the way, thank you for the answer and backstory, I love this kind of details), you know what? Solved it almost instantly. XD I realized where my thinking went wrong: [spoiler] I had counted how many tiles the machine had to move, which is 18, and then proceeded to compute how many loops one could make; this is 3 × 3 × 2, which is 18. And so… I wrongfully deducted I had to have all move instructions at the innermost part of the code, imbricating all three loops, because I totally overlooked the implicit number of move commands (a nice nifty 2). This just goes to show where analytical thinking gone wrong — or rather, with wrong premisses — can get you!! I spent way too much time on this level, this is ridiculous. XD
start with a small piece of code that gets around the first U-shaped bend, and then see what would be necessary to repeat that process around all 3 bends.
You know, this was exactly how I approached the level first, I just kept on having this weird fixation of forgetting to factor in the aforesaid implicit 2 factor. X)
Anyway, thanks for the game! Just my two cents: while the finishing touches are incredible (I would be hard-pressed to come up with this; I still have to take a serious look at Godot, though, and only barely know Bevy when it comes to 3D engines), as you noted yourself, the concept did not match the theme, and was not such a twist (except for the title of that evil ‘twisting curve’!); I am very concept-oriented and could be pleased with literally only geometric primitives but a fantastic concept. Which is what I will (clumsily?) go for in my first attempt at Ludum Dare. Be warned if you stumbled upon it in these parts!
(I don’t want to sound overly critical, it’s just my obsession with originality; the game/prototype was pleasant and commenters already pointed out its merits, I think!)
Oh yes, side question: how did you make the music? This may be useful to me, although I may end up going full mouth-made audio. ;)