Luca Chang

Interested in working on this - #982 is doing some significant refactors to `OAuthClientProvider` that this will need to build on top of.

Does the protocol itself actually matter to models at all? As I see it, the protocol schema is only useful for humans — models are more than capable of figuring...

> Your input/output schemas end up being monstrously large because they can't refer to each other. They're the same size whether the whole protocol or just the input/output schema are...

I see the value, now -- and it's also much more clear that what's being asked for is not what this GH issue is about, and should be moved to...

This seems fairly niche, anecdotally has there been any broader demand for this? IMO web use cases are partially solved by remote servers already, so the most important distinctions in...

> TS SDK could be a target, but ideally anyone building a web based client would be able to partake in this protocol irrespective of what libraries they're using. As...

> [@LucaButBoring](https://github.com/LucaButBoring) I'm getting the interest here do you think this is candidate material for a SEP? > > If so, what steps could I take to help support the...

Also, in the event that this isn't accepted directly, going back to your older comment: > In healthcare standards I've worked on (FHIR and CDS Hooks), we've successfully used approaches...

Left comments, this looks good overall, I just had some feedback on a few minor phrasing points and the diagram.

Relaying here from `#general-sdk-dev`: > There doesn't seem to be a central ticket for all SDK implementations of a particular change, so sharing this here for SEP-1686 -- the following...