Handle uncaught panics

[ribice]

As a Sentry user, I want to capture and report uncaught panics, including stack traces, to Sentry.

This requires an additional process or binary that monitors for these panics and ensures proper reporting to Sentry upon application exit.