# Typed Spin(2,3) candidate-action pilot This directory turns one existing corpus proposal into a small executable gate graph. It is deliberately not a claim that Spin(2,3) is a fundamental symmetry. The candidate is the effective action `S_v1` from document 1199. Its anomaly-free `U(1)_X`, charge-three flavon and BF descent are retained. The pilot then asks whether that action can pass, without prose-only substitutions, through: 1. the four-dimensional action and its declared field types; 2. the `S^1/Z_2` reduction to a 2+1-dimensional level-three BF defect; 3. physical-state, anomaly, ghost and positivity checks; 4. recovery of the chiral Standard Model with nonzero flavour CP violation; 5. a conserved-stress spectral quotient; 6. a measurable two-current/two-probe response. Run the certificate from the repository root: ```powershell node scripts/verify_spin23_candidate_action_pilot.mjs ``` ## Honest present result The candidate passes a small exact/implemented backbone: a typed action is available, the displayed Standard Model plus three right-handed-neutrino charge assignment is anomaly-free, the charge-three compact Higgs phase descends to a level-three BF theory, and a positive conserved two-probe spectral kernel has an `H_2(R)` response geometry. That backbone does **not** connect end to end. The action takes its family projector and neutral mixing maps as input; no physical-state selector derives the observed chiral module; the full Lorentzian ghost/unitarity calculation is absent; no microscopic defect-to-stress operator produces a third response direction; and no Standard Model spectral density has been calculated. The pilot therefore has `physicalAdmission: false` by construction. Missing gates are first-class graph nodes, not comments. ## Branch-discriminating target For a positive two-probe spectral response `C(omega)`, define ```text rho(omega) = det C(omega) / tr(C(omega))^2. ``` Fixed conserved probes in a parity-even Lorentz-invariant vacuum obey an exact two-form-factor plane relation `L(C)=0`; a conformal response collapses further to a ray. The candidate's genuinely discriminating target is therefore not mere positivity or `rho > 0`. It is a reproducible, positivity-preserving third response direction whose plane residual changes with a specified `Z_3` defect holonomy while conventional contact terms and probe changes are held fixed. The action currently does not derive that coupling. Consequently the target is `specified`, not `predicted`. A null result leaves Spin(2,3) as ambient response geometry; a positive result would justify calculating whether an emergent covariance acts on the physical response family. ## Status vocabulary - `implemented`: executable representation or certificate exists. - `exact`: follows algebraically under declared assumptions. - `conditional`: valid only after an explicit upstream assumption is supplied. - `missing`: required evidence or dynamics does not exist in the corpus. No `conditional` or `missing` node is allowed to satisfy an admission gate.