Context
There are various errors such as user errors, system errors, network errors etc. that can happen in Recce. Some errors should be made to behave the same way. For example, non-critical errors might allow for reconciliation to continue with retries, while fatal errors should abort immediately. There is currently no broad error handling mechanism. This leaves error handling to the discretion of the assigned developer.

Expected Outcome
Design and implement a broad error handling mechanism that is extensible for current and future functionality.