| Monitor implementation of what was merely recommended | Obtain the later decision and implementation evidence separately; state unknown status when they are absent. |
| Use an indicator list as the whole arrangement | Name its claim, interpreting owner, timing, receiving use and failure return. |
| Trigger an action automatically from a threshold | Establish the direct operational authority and response rule separately; otherwise return the qualified reconsideration question. |
| Rerun the whole engagement after every new version | Inspect changed claim content and actual reliance; preserve independently usable results and reopen only affected contributors. |
| Reuse an A-specific evaluation for configuration B | Retain the historical A result and obtain B’s exact qualification. |
| Interpret a missed observation as no change | Return the gap and its consequence for reliance, with the responsible receiver. |
| Call later improvement proof that the advice worked | Obtain the direct causal or effectiveness result; chronology and satisfaction alone do not supply it. |
| Keep follow-up alive without an owner or end condition | Establish the actual service or observation obligation, or close with an honest bounded return. |