Practice identifying scales from short musical fragments instead of clean stepwise runs.
This subskill is in beta setup. The generator still needs stricter ambiguity checks before public release.