It's not really usable yet, but I'm working on a fork of Ren'py. I call it Renovel, and development is happening on Codeberg.
The fork starts from the last release of Ren'py before Github Copilot was available to the public. Right now, I'm focused on getting it to run with recent versions of Python. When it's working, the next objective is to replace Ren'py's save system and its rollback feature with Lisien. Then, when you load a saved Renovel game, you'll be able to roll back all the way to the start; and, if you saved while you were rolling back, then when you load, you can also roll forward. Even if you have to roll forward past decision points.
After that, I'm not sure whether it would be better to convert the Dating Sim Engine to run its rules' code in Lisien, or to write my own, similar syntax for expressing Lisien rules in Ren'py scripts. I'll look into it. If you've used the DSE, I welcome your input.
Currently, the Renovel project has the same AI policy as the Godot game engine, which is not ideal, as it permits AI-based autocomplete. Ren'py used Copilot's autocomplete for some years before the rest of Copilot became available, and I don't like that -- both because it made development depend on a third party service, and because that service's training data were neither licensed nor even attributed. To write a good, enforceable policy about it, though, I need to come up with criteria that contributors can understand, which distinguish acceptable autocomplete (such as what JetBrains calls "basic code completion") from unacceptable (like "machine-learning assisted completion ranking"). I am concerned mainly with the source of training data, which is rarely made obvious to users, and which many prospective contributors may have thought little about.
Ideally, there should be an organization of conscientious machine learning developers that pay close attention to the sourcing of their training data, who investigate various code generation options, and list the ones that aren't exploitative in their data collection. Then, I could require contributors to use only the code generation technology on that list. But I don't know of such a list. If you do, please link me to it!