Gravitational Manipulation Theory v1.8: First-Principles Single-Body Screening and Full Physics-Native Constraint Closure

Gravitational Manipulation Theory (GMT) v1.8 completes the transition of the GMT constraint pipeline from legacy-dependent reconstruction to a fully physics-native and reproducible framework. GMT v1.7 removed the legacy grid-generation dependency and reconstructed the two-body mutual response from a native symmetron boundary-value problem. One provenance dependency remained: the single-body screening factor, κ_s, still relied on a frozen bare reference. In v1.8, the single-body screening factor is reconstructed from first principles as κ_s^native = Q_eff^native / Q_bare^native, with both scalar charges obtained from the same native symmetron boundary-value problem, normalization, and tail-extraction procedure. The bare reference is defined by linearizing the identical BVP about the exterior broken vacuum. The reconstruction passes four untuned validation gates: numerical convergence with respect to radial resolution and domain size; recovery of the unscreened limit κ_s → 1; agreement between numerical and analytic bare charges at approximately the 10^-6 level; and a saturated deep-screening law obtained using Aitken extrapolation. The resulting deep-screening relation is κ_∞ = 2.289 μ^1.814, with systematic band C ∈ [2.22, 2.37] and p ∈ [1.79, 1.84]. It agrees with the v1.4 frozen reference to approximately 2–12% across the resolved deep-screening regime. Propagating the native screening surface through the canonical GMT master grid changes the individual experimental exclusions measurably: the Eöt–Wash exclusion count changes from 173 to 155, while the HUST exclusion count changes from 996 to 1007. Despite these changes, the constraint-active Eöt–Wash–HUST overlap changes by only one grid cell, from 39 to 38, with a deterministic systematic envelope of [38, 39]. The independent coverage overlap remains unchanged at 174. Thus, a measurable first-principles change in single-body screening does not produce a material change in the canonical cross-experiment constraint overlap. With the legacy grid, tabulated two-body mutual factor, and frozen single-body κ_s reference all removed, GMT v1.8 closes the remaining physics-provenance dependency of the constraint pipeline. The archive includes the manuscript, all nine ACT I–IX figures, LaTeX source, native single-body BVP and constraint-propagation code, pinned Eöt–Wash and HUST observational inputs, machine-readable results, validation and hardening records, and scripts for reproducing the canonical metrics. All results are conditional on the minimal double-well environmental-potential (W_Ω) class, the published experimental digitizations, band-edge definitions, and stated mapping assumptions. No experimentally demonstrated gravitational manipulation or engineering controllability is claimed.