Durable Repair under Stored-Coordinate Budgets: A Ridge-Trained Checkpoint Separation

We study one-shot factor edits followed by compulsory ridge-regularized minibatch SGD. Capability is the best probability of keeping both task risks safe throughout the continuation window, optimized over the complete legal action class. In a rank-$40$ linear learner at batch sizes near $2,000$, the same stored-coordinate row-write budget gives certain $840$-step survival at a young checkpoint but at most $.001$ survival after old training, on a history event of probability at least $.967$. A task subset of prior mass $7.98$ is certified repairable at editing time; its contribution to the young-minus-old capability gap exceeds $.0768$. More mobility restores reliable survival. The separation holds on an explicit joint neighborhood in cap, ridge, step size, and risk threshold, including caps on both sides of the exact-representation energy. This neighborhood is small and establishes local openness. The proof combines a full-factor lower bound, feasible witnesses, and stopped finite-batch estimates, with executable arithmetic certificates. Seven-history policy experiments support the ridge mechanism; nonlinear age effects are mixed. The result separates age-generated checkpoint states, without matching their predictors or risks or establishing a representation-invariant limitation.