Verify StudioPath fixes and evidence of mean recalculation

Complete verification and deliver the corrected StudioPath application without adding features. Preserve the fixed independent reset seed, literal feedback rendering and draft/evidence separation. Explicitly prove displayed mean recalculation: after Reset select Ari / Sense-making and the Field notes assignment filter, observe No scored evidence; enter score 4 and Save, then observe the displayed competency mean 4.0. Record both actual visible mean values, not only the label or saved input. After the final files, verify all six original acceptance criteria together: selected learner/competency evidence, that before/after mean, unscored display, unchanged complete 144 evidence records/scores around draft creation, actual reload preserving saved score and exact feedback, and executed production tests distinguishing zero/missing and rejecting invalid scores. Recheck Reset→edit/Save→Reset in the same page without intermediate reload; confirm literal feedback containing HTML entities and closing-textarea text survives Save and reload exactly. Keep mobile and desktop browser evidence, execute all tests after the last edit, and report observed results accurately. Change application code only if these checks expose a concrete defect.