Yeah, the design is mostly AI.
Any feedback for the betterment of platform mechanics?
Movement speed is definitely too slow right now - while it makes sense that a Zombie would be slow, it makes platforming a pain in the ass, especially since gravity is relatively strong. This makes the MC's hops basically vertical. They're also relatively short for most platformers, which in addition to throwing some people off is probably going to make the jumps
look easy but
play hard, which is kinda the opposite of what you want to do usually. People like to play games to feel skilled and if the jump looks trivial but they still can't do it, that'll become frustration instead of fun.
Reducing the gravity a bit will help, as might adding mechanics like ledge grabbing or wall-jumping. Increasing the MC's speed will also help. If you haven't, you should also strongly consider allowing the MC to change direction mid-air - once you increase the airtime, it'll be much more noticable if you have or haven't done that, and it adds a lot of control to the player which is usually to the benefit of the game. You can always make the jumps harder, if you need the game to be hard. But the more control the player has, the better.
Another thing I noticed playing the web version for a few minutes is that the hitboxes are a little off. Often I would swing my arms and miss (there ought to be a button for it on the keyboard, not just the mouse, imo), despite looking like I was in range. Then other times I would swing on accident, thinking I was out of range, only to actually hit! I wasn't able to get a clear grasp on how far my attacks went. You could just fiddle with the hitbox, but I think adding a little swoosh effect would also go a long way; then you wouldn't need to worry about the player getting taken out of their immersion by their telekenetic zombie protagonist. Swooshes and dust clouds can cover up a lot of those kinds of rough edges, and give an excuse for the attack to be much longer range - if the enemy contacts the swoosh they get hit!
You might also want to zoom in on the action when the MC is getting raped. It's hard to see, especially if you're not playing full screen. In addition to that, I wasn't able to tell if I could struggle out, which is a problem. Some visual indicator of it being a game over or having a struggle process would be nice; either a "Press E to game over" at the bottom and maybe some darkness at the edge of the screen, or little struggle arrows over a bar like in other rape platformers would do fine. Foreground trees should also be removed, or made transparent enough they don't block the action!
If it is a game over, I'd also prefer it not be, although that wouldn't be too big a deal as long as save points are ubiquitous. The main thing is that I should be able to see a new enemy, go "I want to fuck that thing", and then get the MC railed by it without having to lose a shitload of progress. Not an issue for a demo this small, but it'll become one as the game expands.
Finally, from a pure back-end development standpoint: Write yourself a roadmap, and specify which features you consider core to the game, which are important to the experience, and which are just fluff that sounds cool to add. Then keep track of which updates add what. This'll help fight feature creep and the internet trap of "Guy on the internet has a really cool idea, it's added, but then another guy has a really cool idea and the game stalls as that's added too"