Artian A2 Blind Tabletop Gravity Tests

Three tabletop ways to ask whether gravity carries a source-side ledger signature. This paper develops three blind laboratory protocols for the A2 Law of Endurance: an equal-action/unequal-load interferometer, an echo-cancelled covariance test, and a quantum-source holonomy test. The common object is an A2 address packet whose ordinary GR/QM laboratory shadow is held fixed while the registered source/access ordering is varied. \[ { TA2,r =( OA2,r, P_r, C_r, D_r,q_r, Σ_r,HGR/QM,r,HA2,r, K_r) } \] Version 2.0 makes the experimental verdict fail closed. Every row must pass \[ { GᵉˡⁱᵍⁱᵇˡᵉA2,r =GpacketGregimeGGR/QMGartifact GᵣₑᵥₑᵣₛₐₗGcovarianceGpowerGchronology. } \] The regime certificate must be computed from source history, apparatus geometry, calibration, and frozen noise information without using the target residual. An ineligible or underpowered row cannot falsify A2, and a target-like result cannot promote an ineligible row. The complete predeclared channel vector is retained: \[ { y_r=(Θ_r,V_r,C_r,K_r, s_r). } \] A reproducible secondary signal either invalidates the primary forward model, producing an ineligible packet, or survives it as a qualified cross-channel anomaly. It is neither discarded nor relabelled after unblinding. The paper therefore separates five outcomes: ineligible packet, registered A2 access-map failure, A2 candidate, qualified other/cross-channel anomaly, and ambiguous or underpowered result. \[ { {∂ OA2,r ∂\{ y_r,Σ_r,verdict\}}=0. } \] The protocol design and verdict governance are closed. Apparatus-specific numerical sensitivity, prospective power, and blind laboratory data remain open experimental work. Stable anchors: Concept DOI: 10.5281/zenodo.20772326 Experimental governance: 10.5281/zenodo.21775175 Main book: Quantum Traction Theory: Main Book v10.01 Website: quantumtraction.org Included files: PDF paper, Version 2.0 Reconstruction ZIP containing LaTeX source, style file, metadata, render audit, and SHA-256 manifest