Enqueue webhook handling instead of directly

Multiple callbacks can come in, racing to update state. For some reason they aren't retrying & succeeding, so a workaround would be enqueueing them

Start by locating the webhook callback handling and the state-update path in the repository; the issue does not name specific files or tests. Trace how concurrent callbacks are processed and how retries currently behave. Done means callbacks are enqueued rather than handled directly, avoiding racing state updates and allowing retries to succeed.

Written by the indexing model from the issue text.