Some folks reported issues with this flow earlier today. The reports came in within a single hour period, so I suspect transient network issues.

I spent some time looking at the code here that handles talking to the relay server, and one potential risk here is that if there's an error after the URL redirect occurs, we'll never retry that step. Probably a good idea to reset this, so that if a retry is necessitated after a URL redirect, we can retry the process from the start (including redoing the URL redirect).

Depends on D13580

I tested this one the same as the previous two, which is to say I followed this doc on Notion for both physical iOS and Android devices

Ideally I'd test the isError state here, but it would be really expensive since I'd need to patch @farcaster/auth-kit to force isError, but that package distributes transpiled source