The pilot review combines a late-record query defect, typed source updates and failure recovery. Work from the source population and intended operation before choosing a formula. A fast empty gallery, a green notification or a familiar title can all be misleading when identity or execution scope is wrong.
The tasks revisit earlier boundaries on fresh data. Repair the modeled remote predicate, preserve source-relative amounts, keep denied writes unchanged and explain repeat-request uncertainty. No new remote capability is introduced. The final result should identify what the local app proves and what a real environment still needs to verify. Use the hints after comparing your result with the requested population or saved record.