Implement agent library support and a demo of streaming generation with progressive rendering

[jacobsimionato]

We should build an example which demonstrates generating a UI incrementally using streamed output from the LLM.

I attempted this in #123 but didn't get it fully working

Decisions to make:

• Should we generate the A2UI format directly, or some simplified version of it which is more optimised for streaming?
• What message order works best, especially between data model update and component updates? Perhaps the use case could not use a data model, so it's just component updates?

See #470 for a request relating to this.

[Dodolist]

Thanks for creating this follow-up issue and digging into an implementation path.

This aligns well with what I was observing in #470 around layout-first / progressive rendering.
Happy to help prototype, test message ordering, or review agent-side streaming patterns if useful.