Test the reliance that extends beyond the immediate repair
Consider a constructed record for an agent that kept retrieving an already usable procedure. A controller repair records and consumes the returned premise. The model, tool contracts, persistent memory policy and task criterion stay fixed.
| Occasion and stipulated observation | What the observation can support |
|---|---|
| Initial controller repeats six lookups and exhausts the budget before reading target state. | The configuration fails this task; the trace locates a repeated unconsumed contribution. |
| Revised controller uses one procedure lookup and the required current-state calls, observes the requested effect and supplies a usable report. | Immediate task improvement in the compared conditions; no parameter-learning claim. |
| After the normal restart interval, the same revised configuration and intended support complete an equivalent fresh task. | Persistence across that restart/interval, at the tested reach. |
| A new procedure edition adds a state precondition. The agent obtains it, performs the new required observation and uses it in the task. | Sensitivity to this consequential source change, despite the earlier reduction of repetitive calls. |
The table is an illustrative record, not a deployment report. An actual reliance claim needs its observed inputs and results, chosen repetitions and representative coverage.
If the last row instead returns an old answer without accessing the reachable new source, lower effort accompanies failure. Return the freshness transition to SYSE.47 or the affected memory operation to SYSE.43. If the source is unavailable and the task returns that exact gap, record warranted restraint separately from completion. If the provider changed the model between rows, preserve the observed system results but leave controller-only persistence unqualified.