No External-Field Effect: Cluster-Satellite Lensing, the Cassini Quadrupole and the Outer Solar System Point to an EFE-Blind MOND Sector

Version 1.0 (AI-assisted research programme; not peer reviewed). We test the quadrature MOND law g^2 = g_N^2 + a0 g_N (a0 = (1/2) c sqrt(G rho_Lambda)) against three data sets sensitive to the external-field effect (EFE). Weak lensing of cluster satellites (Sifon et al. 2018, five stellar-mass bins) is fitted by the isolated boost with chi^2 = 0.7 for 5 bins (mean offset +0.01 dex), while including the cluster's field (eta = 0.59-0.81) fails by chi^2 = 99.7, a magnitude-only direction-blind EFE by 99.5, and a no-phantom 'ownership' rule by 131.6. The anisotropic EFE predicts a Cassini quadrupole Q2 = 2.195e-26 s^-2, 6.3 sigma above the measurement; without the EFE, detached trans-Neptunian perihelia are conserved. The data ask for a MOND sector computed from each bound system's own field. This costs the Milky Way's Gaia DR3 outer decline (2.5 sigma) and Chae's EFE signal, and predicts wide binaries boosted as isolated MOND systems (gamma_v ~ 1.20 at 10 kAU, 1.51 at 20 kAU for 1.5 Msun), decided by Gaia DR4. No covariant theory with this property is known; kappa = 1/2 remains fitted.