This is quite good. One thing that was not clear to me until I was at the end of the first chapter was how memories worked. It seemed strange that I needed to be in the first room to form a memory of the picture even though I had it in my inventory and could have in theory looked at it from there. If the rule is that you cannot make memories from things in your inventory, I think you could signal this by auto toggling to the objects in the room tab whenever the REMEMBER action is selected.

Overall, I enjoyed the game.

I think I got to the end of act 2, but there was some text I couldn't scroll through on a smartphone screen, so couldn't get past it. Will have to try again when it isn't 1am!

I grew up on Sierra and LucasArts games and wanted to build something in that mold, but about the questions I actually think about now: what intelligence is, whether prediction equals understanding, what a mind needs beyond information. So the puzzles are philosophy of mind, played straight. You operate Searle’s Chinese Room from the inside — following the rulebook, understanding nothing, until a card arrives that no rule covers. Mary’s Room is a location: a colorless room containing complete knowledge of red, and the solution requires physically carrying something out of it. The game has a REMEMBER verb, and memories are inventory items — the final door won’t open for a player carrying facts but no weight. The frame story is Devs-flavored: a machine called LAPLACE that has already printed a transcript of everything you’re about to do, built by a man who wanted to watch his drowned daughter one more time. Free, browser-based, works on phones, ~45–60 minutes, two acts, five endings. EGA palette, verb buttons, deliberate deaths. Honest disclosure: I designed and wrote this in collaboration with Claude — a game about machine cognition, co-built with a machine. That felt less like a shortcut and more like part of the experiment. Happy to answer questions about the process or the design. https://schwarzarno.itch.io/essel