Integumentary Barrier Seam: Live-Verified, Identity Holds as Closed-Form

The skin volume arrived, so the epidermal barrier seam was tested live. Substrate drift is 0, the epidermal barrier is real (TP63, KRT14), and the skin volume independently declares an immune out-seam — a reciprocal handshake. The epidermal-tolerance identity stays immune-owned closed-form: confirmed-real-surface, not contradicted, but not engine-verified, since the skin defers immune tolerance. Grade [V].

The second live cross-package verification against a real sibling — and a third distinct outcome. With integumentary_vp_site v1.0.0 (concept DOI 10.5281/zenodo.20754541) on disk, the §18 harness measures cross-volume substrate drift = 0.0: the epidermal barrier and the immune tolerance machinery share one byte-identical R19 substrate, and the barrier surface the candidate posited is real (epidermis TP63, keratinocyte KRT14). The skin volume independently declares a reciprocal immune out-seam (out__immune_hematologic__urticaria) — a bidirectional handshake. But the skin volume models the barrier structure and defers the immune tolerance switch as an out-seam, so the epidermal-tolerance identity (induction = epidermal-antigen + spinodal(1.0)) remains immune-owned closed-form: confirmed-real-surface, not contradicted, not engine-verified. Between the gut (verified identity) and the marrow niche (retired identity), the skin is the honest middle. Engine hash byte-identical.

The second live test — and a third kind of answer

The marrow-niche seam (chapter 20) retired its identity; the gut seam (chapter 16) verified one. The skin barrier seam lands between them. With integumentary_vp_site v1.0.0 on disk, the chapter-18/19 epidermal barrier candidate could be tested against the live skin engine — and the result is neither a clean promotion nor a retirement, but an honestly partial outcome, reported exactly as measured.

The barrier surface is real and shares the substrate: drift 0

The harness loads the skin engine and compares its R19 spinodal 2(γ/3)^1.5 and barrier γ²/4 against this volume's, bit-for-bit, over γ∈{1.0, 1.3225, 1.4892}. The drift is exactly 0.0 — the epidermal barrier and the immune tolerance machinery sit on one byte-identical substrate. The barrier surface the v0.18.0 candidate posited is real: the skin volume builds the epidermis (TP63, spinodal 0.61335644) and the keratinocyte barrier (KRT14, spinodal 0.69962471) on that shared substrate. So there genuinely is an epidermal barrier surface for the immune tolerance switch to localise to.

A reciprocal handshake: the skin volume declares an immune seam back

The immune package declared a skin barrier-immunity candidate (immune→skin) in v0.18.0. The skin volume, written independently, declares an immune out-seam in the other direction: its own seam manifest carries out__immune_hematologic__urticaria, a contract to immune_hematologic for the mast-cell / histamine effector (urticaria / angioedema). Both volumes recognise the same barrier↔immune adjacency from opposite sides, with no coordination — a bidirectional cross-reference that can now be closed. (The skin emphasises the mast-cell effector; the immune-side candidate emphasises epidermal tolerance — related immune functions at the same barrier surface.)

The identity stays immune-owned closed-form — not verified, not contradicted

The stronger claim was a shared-substrate identity: that the skin's epidermal tolerance threshold equals epidermal-antigen + spinodal(1.0) in closed form, the offset invariant at ±0.38490018. The live engine neither confirms nor denies it, for a clean reason: the skin volume does not implement an immune tolerance switch at all. It models the barrier structure, physical and thermal insult tolerance, pigment, appendages, and — notably — autoantibody-driven adhesion saddle-nodes (pemphigus and pemphigoid, where an antibody drives a desmosomal or hemidesmosomal bond below its spinodal, with hysteresis). That last is the same R19 saddle-node formalism applied to a different compartment, driven by an immune effector from outside. But the immune tolerance decision itself the skin names as an un-modelled out-seam. So there is no epidermal tolerance threshold in the skin engine to equate. The identity therefore remains the immune-owned closed-form result of barrier-surface agnosticism — now with a confirmed real surface and a reciprocal seam, neither over-claimed as engine-verified nor retired. The absolute epidermal-antigen scale stays skin-owned [O].

Why three different outcomes is the discipline working

Three sibling volumes have now been tested live, and they gave three different answers: the gut identity verified (the digestive course reproduces the saddle-node), the marrow-niche identity retired (the niche is built by a different switch), and the skin identity stands as immune-owned closed-form (the barrier surface is real and reciprocally recognised, but the tolerance switch is immune-owned). A framework that returned the same verdict every time would be asserting, not testing. Recording each outcome as it actually came back — verified, retired, or closed-form-not-contradicted — is what keeps the cross-package map faithful.

What this is — and is not

This chapter adds one live verification and one reciprocal-seam handshake; it adds no measurement to the thirty-five stress targets and changes no number in the engine. The cross-package statement is a direction-and-structure result — direction/class only, not medical advice, and not a validation of VP theory. The immune package still reproduces alone: with the sibling absent the harness SKIPs cleanly and the immune-side contract is byte-identical with or without it, and the engine emit() hash stays byte-identical. The seam layer and the harness each carry their own 2×sha256 (seam 9357f23b…, harness 2b07d4d5…), separate from the engine hash.