A bit more context on why I built this:

I like MPA simplicity, but I also wanted progressive in-page updates without turning everything into a SPA with client state, router logic, and hydration complexity.

So this project is basically me asking:

Can the server remain the source of truth for UI state, while the client only handles DOM projection for named slots?

I’m not claiming this replaces React/Next/etc. I’m trying to see whether this model is useful for a narrower class of apps: - streaming interfaces - server-driven workflows - multi-step forms - dashboards - agent / assistant-style UIs

Very interested in where people think this idea is promising vs fundamentally awkward.