Einstein-Karahan Framework v129 — THE TWIN-SCALAR GRAVITY SIGNATURE: Exact Two-Pole Predictions, Poisson-Null Source Engineering, and a Blinded Underground Search Architecture

This work presents Einstein-Karahan Framework v129, a major post-v128 update centered on a falsifiable twin-scalar gravitational signature, a Poisson-null / Yukawa-active source architecture, and a blinded underground search strategy. The corrected Common-Vacuum branch contains two coupled scalar eigenmodes on the stable connected interval 1 < s < (3 + sqrt(17))/4. Within this branch, the linear matter residues satisfy the exact sum rule eta_- + eta_+ = 1/3, the light residue eta_- is a strictly monotonic branch coordinate, the heavy residue remains bounded away from zero, 0.327683... < eta_+ < 1/3, and the scalar mass hierarchy obeys m_+ / m_- >= 19.97818.... These relations create a direct no-refit falsification structure: a measurement of one residue fixes the branch coordinate, which fixes the second residue and the scalar mass ratio; after one absolute scalar mass or range is calibrated, the second scalar mass or range becomes a prediction rather than an additional fit parameter. The second major development is a Poisson-null / Yukawa-active (PNYA) source architecture. In the ideal continuum construction, the density perturbation is generated as delta rho = nabla^2 f, causing the exterior Newtonian potential of the perturbation to vanish while a massive Yukawa response generically remains nonzero. A C6 complement symmetry further restricts the ideal angular density content to m = 3 mod 6 and removes the m = 0, 1, and 2 sectors. The practical realization uses binary W/Si source engineering, curved-cell moment control, segmented radial trim, and a blind empirical Rank-5 Newton transfer calibration. Manufacturing connectivity, real actuation, and metrology remain explicit SOURCE_GAP items. The third development is a blinded long-channel underground search architecture. The current design uses a 500–1500 micrometre measurement window and a blind 13–18 Hz frequency corridor, with 13 Hz / 260 rpm as the present conditional Pareto commissioning point under the stated Q and passive-settling assumptions. Environmental gravity, mechanical transfer, and Rank-5 calibration are required to share one preregistered weak-signal systematic budget. Candidate frequencies are selected only from environmental and null-channel data, with simultaneous multi-frequency confidence control and an independent holdout survey before science unblinding. This release is theoretical and computational and defines a prospective experimental program. No Twin-Scalar particle has been experimentally detected, no new fundamental interaction is claimed from data, and no natural-world detection is reported.