Char p Frobenius: The Unifier in Characteristic p

This paper tests the Unifier in characteristic p. In 2026a, we proved no unital map T → E for H = S^1. In 2026b, we generalized to the triple H^n(T(H)) = (dim H^n, dim C^n, 1) for all finite H, all n. In 2026c, we defined the Bridge Conjecture. In 2026d, we proved the eraser must be CPTP decoherence. In 2026e, we proved all four follow from one Unifier Axiom: [x,y] = 1, and no unital φ: T(H) → E preserves [x,y]. In 2026f, we categorified H: histories as objects, T as functor, erasure as functor, No-Go as natural. Here we ask: does the Unifier survive in characteristic p? In characteristic 0, [x,y] = 1 is sharp. In characteristic p, the Weyl algebra A_1 has a large center, the Frobenius map F: a ↦ a^p is central, and x^p, y^p commute with everything. We prove the No-Go survives, but in a new form: erasure cannot kill the Frobenius center. We define the char p triple (dim H^n, dim C^n, 1_p) where 1_p is the Frobenius obstruction, and prove it cannot be erased by any unital map. This does not change the experiment. It changes the algebra in which the experiment is interpreted. If the Unifier failed in char p, it would be a characteristic 0 accident. It does not fail. Tseglin 2026g = Ch. 142. Cites Tseglin 2026a DOI 10.5281/zenodo.23977165, 2026b DOI 10.5281/zenodo.23980498, 2026c DOI 10.5281/zenodo.23980738, 2026d DOI 10.5281/zenodo.23981860, 2026e DOI 10.5281/zenodo.23982063, 2026f DOI 10.5281/zenodo.23982359.