Transformation-Stable Evidence Interoperability (TSEI): Specification v0, Extraction Boundary, and One Independently Grounded Instance

Transformation-Stable Evidence Interoperability (TSEI) is a producer-neutral mechanism for deciding whether a declared transformation preserves the semantic surfaces that a domain profile marks as protected. It does so by recomputing source and target evaluations rather than trusting either artifact's self-report. The specification defines a five-outcome verdict model; edgewise cycle evaluation that prevents later endpoint restoration from erasing an intermediate failure; closed-world coverage over the structural paths observed in one evaluated pair; deterministic selector precedence; an explicit normalizer-authority trust root; and canonical comparator requirements. This publication has three distinct evidence layers. First, Sections 1-16 state the generic normative mechanism. Second, an extraction boundary is mechanically exercised on a synthetic non-domain object, with runtime import closure excluding ReceiptOS and Chronicle domain terms; the current repository location is therefore a packaging fact, not a dependency of the generic core. Third, a separately governed Protocol v2 case study records one post-ratification independently grounded instance. In that instance, a prior collaborator acting as Authority derived attribution sets from answer-free Object A before oracle disclosure; the originator's commitment was opened from the exact E0 artifact at rest; and the independently frozen Object B agreed with the revealed oracle over the complete 12-case universe. The case study establishes bounded artifact grounding for one instance. It does not independently machine-verify the human semantic judgment, does not establish a complete real-run claim, and does not provide cross-authority replication. The Authority relationship is disclosed as EXTERNAL_PRIOR_PROTOCOL_EXPOSURE: operational information separation was enforced, but organizational or stranger independence is not claimed.