nx_engorge_gate.nx

nx_engorge_gate.nx -- TISSUE SWELLS BECAUSE ITS REST STATE GREW (softbody SB16, 2026-08-25). THE CROSS-BOARD CONTRACT THIS CLOSES. procgen's arousal row names its own blocker: "a pure normal push is not volume-conserving, which the conservation gate above would refuse." That gate is ours, and it is right to refuse -- displacing surface vertices leaves every rest length and rest volume untouched, so the constraints are all violated and the next substep pulls the tissue back. The tissue never swells; it fights. Growing the REST STATE instead lets the solver find the new equilibrium with every constraint satisfied the whole way. THE ORACLE IS GEOMETRY, NOT AN OPINION. Isotropic linear growth by s must scale measured volume by s^3. That is a closed-form prediction, so this gate does not ask whether the tissue "looks bigger" -- it asks whether the measured volume ratio matches the cube of the commanded factor. A model that merely pushed the surface outward would move the skin without tracking s^3. ⚠ SCOPE: the MECHANISM only. How much any real tissue engorges, and how fast, is a measurement nobody in this estate has mirrored, so the factor is the caller's and no band is invented here -- the same cite-or-abstain split the friction coefficient ships under. license_tier: ORIGINAL No hw writes (Rule 26). expect_exit: 0