The Representational Jurisdiction Objection: The World A / A+B Frontier

This record consolidates the 2026 Origas programme on representational jurisdiction and conservative reconstruction. It identifies an exact, task-relative boundary between what a declared readout P can reconstruct and what remains unresolved on its fibres. The central criterion is: J∈Jur(P) if and only if J is constant on every relevant fibre of P. When two source states satisfy P(x₁) = P(x₂) while a task requires J(x₁) ≠ J(x₂), the task lies outside the jurisdiction of that readout. The coherent options are then to prove factorization through P, restrict the claim, or add a genuinely source-sensitive conservative channel. The record develops this boundary through bounded conservative completion A+B=(P,K,T), information-processing closure, conditional information gain, reversible four-sector coding, Point-0 non-collapse, bridge geometry K, Schur-inertia sensitivity, order-sensitive reconstruction, provenance/history, and decision-theoretic limits of P-only policies. A comparative scientific annex places these contributions alongside cited work by Hong Wang / Wang-Zahl, Alain Connes, Zhou / OpenAI, and Zeta23. The comparison does not claim to replace their domain-specific results. It identifies a distinct bounded question: whether an exact representation has sufficient information for an external task. This record claims priority, to the extent documented by the cited timestamped DOI records, for the Origas formulation and cumulative synthesis of task-relative representational jurisdiction, bounded conservative completion, and the World A / A+B reconstruction framework. Physical interpretations are explicitly separated from the exact mathematical layer. B_math ≠ B_phys. RingCross, literal physical +t/−t, and a universal T_phys remain OPEN pending independent, preregistered and reproducible empirical discrimination. The record contains three files: the main formal publication, an English comparative scientific annex, and a machine-readable table of the ten exact bounds.