Certified Export Without a Global Hypersurface: World-Tube Boundaries, Record Fluxes, and Covariant Export Accounting

Which heat, which records, which exports belong to an account? The fifteenth grouping of this series (Papers 51-53) closed its thermodynamic ledger honestly — types, qualification, accounting — but in a deliberately fixed-slicing dialect: the certified-export heat delivered per unit of one laboratory time, "the declared window" a coordinate-time interval, the eligibility gate evaluated in the device rest frame. Paper 53's own closing sentence hands the residue forward: the sixteenth grouping's covariant setting, where "the declared window" itself becomes the question. The presumed authority under audit here is the Global Hypersurface — the assumption that a single global simultaneity cut fixes what an account counts. It is the accounting-side cousin of Paper 45's Global Simultaneity, and the seats are distinct: 45 audited the kernel's dynamics; a fully covariant dynamics can still be accounted through a preferred slicing, and Papers 51-53, by declaration, did exactly that. This paper — the fifty-fourth, opening panel of the Covariant Completion Program (Papers 54-56), whose charter it defines — audits whether the certified-export account can be carried by the apparatus world tube instead. Delivered: a covariant window triple (world tube, closed proper-time interval, record region), with ledger membership by tube label; a certified-export current — an architecture-level bookkeeping current at exactly Paper 53's row-attribution grade, selected from Paper 45's pathwise per-event export stress tensors by Paper 42's chain-membership certificate, an accounting selector on the completed event ledger, never a dynamical source selector — inheriting conservation, causal support, and monopole-freeness summand by summand; a flux normal form delivered at two typed levels — a pathwise ledger-flux identity (the realized certified sum equals the current's lateral-boundary flux, term by term, with per-event value fixed by Gauss's theorem under Paper 53's consumed zero-rows, and with straddle and weak-field remainders displayed) and an intensity-level per-unit-proper-time form at the type of 53's own normal form (dQₑxp/dtau = hbar*omegac*nu) ; a verdict-invariance theorem — admitted, refused, and zero-export consume intensity-level and record-level data only and are constant over the declared class of admissible re-descriptions, with the one live smuggle candidate (the qualification gate's stationarity clock) caught and re-typed to tube proper time; and a rest-frame recovery lemma, exact at the intensity level — Paper 53's per-event and per-unit-time forms and Paper 42's audit identity nu = etacert * kappa * nᵣes return form to form, frozen corner numbers intact (certified intensities 2. 28e3 and 5. 32e4 per second; export fluxes 9. 1e-21 and 2. 1e-19 W), with every kinematic correction displayed and orders below tolerance (time dilation 5e-17; Unruh scale 4e-20 K against the licensed 73 mK). The Planck-Ott controversy — the sixty-year dispute over whether a boosted body's heat is Q/gamma or gamma*Q — is imported as the literature display of what slice-anchored heat accounting does in public, and dissolved for this account by construction: the normal form books a scalar flux along a declared tube field and never asks which frame's energy. The realized ceiling is scoped, not certified, and the abstract says so: the paper closes the fixed-background, weak-field, tube-accounting conjuncts (b1) - (b4) of its declared register, while (b5) — a boundary definition that stands without a declared fixed background — remains the named frontier (Paper 46's general region-algebra construction and its beyond-weak-field row), scored open here and built elsewhere. Verdict, from the charter-frozen domain: CovStatus (boundary) = scoped — at model grade, on the declared tube, window, and sector classes, conditional on the adopted kernel, Paper 45's certified core, and its minimal dressing assumption. No new physics is adopted; no calibration is spent; no datum is scored; Paper 53's Clausius pass is not re-scored on any branch; and no Global Hypersurface is consulted anywhere on the way. The Hypersurface is never refuted; it is unemployed: a window triple, a per-event scalar against a declared tube field, an outward co-normal — and nothing left for a preferred slicing to decide.