Great art and i liked the simple music you made too :)
I liked the concept of rewinding time with foresight to fix things before they happened too.
Ultimately, I agree entirely with what SeminiGames said below; great art, animation, sound, concept; clarity/intuitiveness could be improved upon, as could execution of the theme.
So overall, I thought it was a cool game and concept, just some rough edges :)
Areas of potential improvement:
>More restart points (or a different way to reset/rewind the situation); being able to reset time to about 5 seconds before every scenario would have been much nicer than having to start at the beginning.
>More intuitive hints of what is interactable. Alas, I also got stuck figuring out what to do (not that I can complain too much, my game is also unintuitive haha). At one point I avoided the agent (so that he wouldnt follow me so i could have more time) and walked around the whole level to see if i could find another 5G tower, and since most of the buildings weren't interact-able by the time I got to the shop I just assumed it wasn't interact-able either. Fortunately I came back here and found your comments and figured that part out. But even after reading your comments I couldn't figure out the acid part, i couldn't put the dog treat on the manhole cover because the dog kills the agent before he gets that far, and from what i could tell i couldnt buy multiple of them and lead the dog on a trail to it. That's a long winded way of saying 'give me more hints and i'll happily finish playing your game' :P A potential way to improve in the future would just be a an outline or colour near anything you can touch, which would allow the player to know what to try and what to not bother trying without adding much dev time for you.
>Potentially dont make us have to keep the agent in sight. Since I have lived the future and know what is going to happen to the agent I would think I could leave him while I go and fix the future to work for us :P
Although I'm not sure if this advice is helpful or not, since this is probably stuff you were going to do and just ran out of time (like me haha) during the tight timelines. So instead my advice for next time (hopefully you make another game-jam game cause this was definitely cool!) is to focus on one mechanic that you can repeat a couple of times and use that in creative ways. So for example, rather than being able to push things AND enter a store AND get objects AND drop objects, maybe just allow the character to push things; and then have scenarios where you have to push the man to stop the car; push the dog out of sight/push a cat in its way; push a box to block the manhole. That way its less effort for you, and the player has less abstract thinking to do, they can just focus on 'okay, how can i use my push skill to solve this new problem?'. I think the thing i've learnt from this game-jam is scope management is the most important thing for success in a limited time setting.
Sorry for the super long post, hopefully its helpful in some way. And if it sounds too negative I didn't mean it to be, it's just that i see quite a bit of potential in your idea and liked the work you put in :)