Admissibility Is Not Probability: A No-Go for Closure-Only Probability Claims and an Audit Taxonomy for Candidate-History Frameworks

This public claim-control working paper studies candidate-history proposals that first define an admissible set of histories by a closure, consistency, or admissibility predicate. It asks what, if anything, follows probabilistically from that admissible set alone. In the terminology of the prior B1/S0 guardrails, the answer is negative: an admissibility layer is not yet a probability layer. B2 is therefore a no-go and audit-taxonomy paper for closure-only probability claims. It does not derive the Born rule, define a physical selector, solve Bell/no-signaling, supply an actualization rule, prove empirical equivalence to standard quantum mechanics, validate RRCF, Complete-History Selection, or any candidate-history framework, or propose an experimental residual. Its limited purpose is to state what an admissible set of candidate histories does not yet license, and to classify which additional probability-layer structures a later positive proposal would have to declare. The central negative claim is that admissibility is not probability. If a proposal defines an admissible class of candidate histories, then it has supplied an admissibility layer, not a probability layer. Any later probability claim needs an additional declared probability object, for example a countably additive baseline measure, a conditioning rule, a weighting principle with reference measure, a stochastic kernel, or an operational map, together with separate Born/no-signaling/no-hidden-postselection checks over the final operational law. B2 records four modest no-go claims: an admissible set alone does not determine a probability measure; conditioning on admissibility is not a derivation of probability; unconstrained reweighting is not automatically no-signaling; and sector-internal record consistency does not determine external sector weights. A short RRCF special-case audit template is included only to state the corresponding probability-gap result for record-reconstruction closure. It is not a construction of a physical RRCF criterion and does not supply a probability law, selector, actualization rule, Born derivation, GR/QFT unification, or prediction.This is Paper B2 v0.2e in the public guardrail chain A -> D1 -> B1 -> C0 -> S0 -> B2. It is a public claim-control working paper and no-go/audit-taxonomy note. It is not a positive probability theory and does not validate RRCF, Complete-History Selection, or any candidate-history framework.