Thank you for the detailed feedback! This kind of feedback is what will help me improve in the future. I know that AI was heavily used in the development of the game as I explicitly said so. But rest assured that all the decisions made were mine and not AI running in autopilot. This means that I might've made some poor game design choices and that's why your feedback helps. The core of the game is built around chaos hence the name. I didn't want it to be structured or uniform so that I follow the luck based pattern. Yes a lot of the suggestions were made up by AI agents, but it was still me steering and controlling it. I hope you were able to tell that I stayed in the loop myself because I tested and architectured everything (scripts layout, scriptable game objects, what the hierarchy should look like, what should be tunable for the inspector, ..etc as I am a software engineer afterall and I know how to scale a codebase) upto my abilities. Yes agents can do tests and what not but those aren't enough. I tried to jam pack a lot of luck based features and that might've made it confusing as to what the goal is in terms of the gameplay. My goal in general was to have fun regardless. Like some players might want to complete a run as fast as possible while others might simply want to survive and experience the chaos. For me, I view AI as a tool that if wielded correctly can help developers expand the scope and achieve much more in less time. Maybe my design approach for this game was weak, but nonetheless I enjoyed the process and will continue to improve!
I did learn a lot this jam. Like players generally will only give each game around 5mins so adding unlockables and shop mechanics might hurt the game and be wasteful. That time could've gone into polishing other features that the player will encounter in the first run even more like the main character as you mentioned.