Just wondering, why did you decide to build this around proprietary, managed cloud infrastructure when the core utilities are open source? Doesn't this unnecessarily lock many use cases into AWS?
Only recently have I started the process of formalizing the framework as it has existed, so there hasn't been any work done on supporting a cloudless implementation. It's on the radar (I have a notice on the top of the git readme about this very topic and needing contributors).
If you are interested in this, feel free to join the discord posted on the github. I'm very much open to discussion and dialog.