A finite flare creates an early allowed branch in the 2024 eRASSt J0456-20 clock test

We test whether the 2024 non-detection of a forecast flare in eRASSt J0456-20 excludes a period change proportional to flare fluence. We reuse signed cycle times, fluences and aperture-count products. We add a finite flare duration and compare onset and floor-crossing clocks for a 32-day rise and a grid of amplitudes and plateau lengths. The floor-crossing marker has no physical solution when the flare peak is below the faint-state floor. The published half-amplitude forecast assigns about 31 percent probability to that regime. A finite flare also permits an early, completed event, so a non-detection need not give a one-sided timing limit. We give aggregate fit results and a general evidence bound for a shared probability of a dark passage. This is a sensitivity study with fixed background, floor and instrument calibration. It does not infer a dark-passage probability, establish an orbital clock or provide new observations.