A crash left our housing verification request marked pending locally, although it may already have reached the FR10; startup sends every pending row, so a word that once meant 'not sent' is now authorising another attempt at work we cannot account for.
Automatic resubmission is disabled while we sort it out; the developer can reproduce a crash after sending but before saving the reply, and that row now reopens as uncertain.
Can the operator see the housing and attempt identifiers on that uncertain row? Our caller once had a good description of a wait and no way to identify which job needed help.
Naomi, the proposal uses the retained result tied to the original attempt plus an authorised review when that cannot settle it; Daniel, those identifiers are on the revised screen, not just hidden in the log.
The offline cases now cover matching retained result, no result, and reopening after a saved review decision; no-result remains held and none silently resubmit. Our operator walkthrough found the call contact unclear, so that part still needs finishing.
Have the actual contact receive the example call once named. It is an easy way to discover which identifier they need before your first real recovery call arrives.