As part of my work on #267 (closed), what has been left out at the moment was auditing the codebase for all those occurrences of panic and error which would reflect in a poor/no error on the frontend side.

We should audit these occurrences and at the very least have a top-level exception handler to turn them into a proper error, but even better we should convert those into a BackendInternalError so that they have a proper translation into a FrontendError.