A two-stage phenomenological framework for testing shared directional signals in cosmological anomaly claims

This working paper proposes a two-stage, data-first framework for adjudicating directional claims across cosmological anomaly datasets. Stage 1 defines a phenomenological shared-axis testing methodology using public probes such as Type Ia supernovae, quasar number-count dipoles, peculiar-velocity catalogues, CMB large-angle consistency checks and BAO constraints. Stage 2 is explicitly deferred: only if a statistically meaningful shared or scale-dependent directional signal is found should microphysical interpretations such as baryon-asymmetry gradients, anisotropic dark energy or compensated long modes be developed. The paper is intended as a methods framework and research proposal rather than a claim of discovery. Its central contribution is the statistical infrastructure for comparing isotropic, shared-axis, probe-specific and scale-dependent directional models using nested likelihoods, information criteria and falsification conditions. Negative or misaligned results are treated as scientifically informative constraints on any future directional-bias theory.