The CMB sky shows six statistically significant large-scale anomalies unexplained by LambdaCDM. The quadrupole (l = 2) and octopole (l = 3) are mutually aligned along the ecliptic plane — the ‘Axis of Evil’ — with p < 0.1%. The quadrupole is anomalously low in power (5-10x below LambdaCDM). The CMB shows ~7% hemispherical asymmetry. A ~5-degree Cold Spot in the southern hemisphere is colder than LambdaCDM produces (p < 0.01). Parity asymmetry distinguishes even- and odd-l multipoles. An intrinsic dipole of order 10^-4 is expected but not yet measured after kinematic subtraction. Under LambdaCDM these six anomalies are treated individually, each dismissible as a ‘look-elsewhere’ artifact. The strongest statistical signature — that the anomaly axes align with one another — is not explained. We show that in the Three Time Dimensions (3+3) spacetime framework (de Haan 2026, book DOI 10.5281/zenodo.19633127), the third time dimension t_3 is compactified as a discrete two-sphere S^2 with 2^152 Planck-area cells, and the six CMB anomalies are six projections of a single S^2 azimuthal structure onto the last-scattering surface. The structure has four components, all sharing the reference direction phi = 0 degrees: hot axis (cos phi, drives dipole and quadrupole), trisection (cos 3phi, drives octopole), 12 pentagonal defects (cos 12phi, drives Cold Spot and 11 additional spots at 30-degree spacings), and CP tilt arcsin(1/4) = 14.48 degrees (drives hemispherical asymmetry). These are not four independent parameters but one geometric fact seen through different angular-scale projections. The framework makes three firm near-term commitments. (1) All six anomaly axes must converge on one direction — the trisection axis (1,1,1)/sqrt(3) set at the Bang. Planck 2018 observes convergence at 98-99.97% CL; CMB-S4, Simons Observatory, and LiteBIRD will test decisively through the late 2020s and 2030s. (2) The Cold Spot is not unique: eleven additional cold spots must exist at 30-degree azimuthal spacings around the defect ring. (3) An intrinsic CMB dipole of order 10^-4 survives after kinematic subtraction, aligned with the trisection axis; DESI peculiar-velocity surveys expected 2028 should detect it. The paper extends beyond the CMB. The same trisection axis should align with the Great Attractor direction and with the preferred axis of galaxy spin alignments. Large-scale matter distribution should show 3-fold azimuthal symmetry at sub-percent level, detectable by DESI DR2+ and Euclid. DES, KiDS, HSC surveys already show intrinsic alignments > LambdaCDM at > 10 Mpc, consistent with the CMB preferred axis. The same S^2 azimuthal structure that produces the particle spectrum and the fine structure constant also stamps its pattern on the entire large-scale universe. The CMB is not a random background; it is a photograph of the S^2. The paper is honest about statistical weakness. Individual anomaly p-values (0.1-2%) are not individually decisive — the look-elsewhere critique applies. The framework’s strength is the joint alignment claim: the probability that six independent anomalies share one axis by chance is astronomically small, and this joint prediction is made before the measurement. Section 9 presents the honest status, acknowledging what a skeptic would require and what the framework can deliver.