It was done just great, amazing. So many functions, features, convenient operation. It all comes down to the shortcomings of the neural network. When disabling various features, the memory is reset... So, for example, I used "total control", did a couple of actions, wanted to see the character's reaction without my control, disabled this feature and the dialogue just returned to the very beginning.
Literally:
"NARRATOR: <<............ She looks at the man, her red eyes wide, and whispers, 'What... happened?'>>
(I have disabled all enabled features here)
USER: You are unable to directly communicate with your host. They act without your input.
NARRATOR: Rumia, drawn by the fading warmth of the dying bonfire, approaches the campsite, her darkness sphere casting eerie shadows on the tents. The travelers, deep in slumber, remain oblivious to her presence. As she nears a tent, her simple mind fixates on the faint scent of human sweat and fear emanating from within............."
But I think it all depends on the capabilities of the neural network and the memory of the free model or something. Although, maybe I'm silly and I don't understand something... But I expected that disabling the features would allow the story to continue on its own, without my participation, and to watch the events unfold when the character finds himself in a stalemate after disabling my control. You, in turn, have done a herculean job of setting it up, for which thank you very much.
But otherwise it's just super! It's brilliantly done)
I really liked it.
Sorry if I didn't describe the problem clearly enough, unfortunately English is not my native language, I use a translator... So yes.