Source sums routine forty-five, verification twenty-five, scoping fifteen and architecture fifteen to one hundred, then divides each by the total. Bars compare routine and verification shares at ten pixels per percentage point. They are entered task assumptions, not measured workforce shares. General vulnerability arithmetic weights routine fully, verification point four and architecture point one, giving fifty-six point five percent for these inputs. But the exact software default overrides that with assigned sixty-eight point five. Bars compare formula and fixture at eight pixels per point. The explicit branch is not a validated probability of automation or confirmation of the study attribution. Source orchestration toggle adds twelve percentage points to junior delta and verification toggle adds eight. On the stored minus-nineteen baseline their arithmetic sum produces plus one when both are enabled with unchanged task shares. Bars compare the two increments at thirty-five pixels per point. These constants do not establish actual employment effects or career outcomes. Original complete native JSON and reset/resize/first-footer evidence remain separate from model validity.