Stage 19 Observable-Projection Audit for the HLV Framework: OP1–OP6 Synthetic Forward-Projection, Backward-Reconstruction, Projection-Frame, and Frame-Equivariance Guard Runs/Independent Clean-Room Replication of the Stage 19 OP6 Observable-Projection Bridge in the HLV Framework: Seal-3 Result Package

This dataset contains the independent clean-room replication package for the Stage 19 OP6 observable-projection bridge audit in the Helix-Light-Vortex (HLV) framework. The replication was performed from the code-free Stage 19 OP6-SPEC v1.0 and the R1-SPEC v1.3b documents only. Marcel Krüger's Stage 19 Python implementation was not used as an implementation source. The purpose of the replication was to test whether the OP6 synthetic projection-frame and frame-equivariance guarded observable-projection identification bridge can be reproduced under an independent code path. The result is OP6_STRICT_PASS. The replication used R = 2, rho = 0.90, 64 fresh sealed seeds, one target model, and eight primary null families. The target-nearest fraction was 63/64 = 0.984375. The pooled median margin was 16.4248 with pooled sign-flip p = 0.0005. All per-null median target margins were positive, with Bonferroni-corrected p = 0.0040 for every null family. The false-target-nearest fraction was 0.0 for all eight null families, including the previously critical edge_orientation_shuffle and permuted_projection controls. The package includes the independent scoring source, Seal-2 and Seal-3 manifests, sealed matrix files, raw per-sample classification tables, distance-component breakdowns, candidate features, signed-axis histograms, target margin tables, pairwise margin summaries, false-target-rate tables, summary-by-model tables, and evaluation JSON. The admissible claim is deliberately narrow. This dataset supports only an independent-code-path reproduction of the OP6 synthetic observable-projection identification bridge under the declared finite computational protocol. It does not validate empirical observables, physical spacetime emergence, continuum quantum field theory, particle spectra, mass generation, Standard-Model recovery, BD17A, theta = 2/9, C3/A5 confirmation, or HLV ontology. The result strengthens the Stage 19 synthetic bridge-audit layer and provides an independently reproduced OP6 Seal-3 result package for later manuscript integration and claim-state governance. This archives contains the Stage 19 observable-projection audit package for the Helix-Light-Vortex (HLV) framework, covering OP1 through OP6. The purpose of Stage 19 is to test a restricted post-R1 bridge question: whether the finite native-6D orientation-sensitive carrier fingerprint, previously reproduced under the R1 clean-room replication protocol, can remain identifiable after synthetic observable projection, measurement degradation, backward reconstruction, and null-model comparison. The archive includes six sequential runs: OP1 tested a first synthetic observable-projection and backward-reconstruction pilot. It produced a strong pooled reconstruction advantage, but the strict pairwise null ladder remained blocked. OP2 tested observable-projection identification margins. All target-vs-null margins passed, but a false-target-nearest control blocked the strict claim. OP3 added an orientation-consistency guard. This removed the edge-orientation-shuffle false-target mode, but a permuted-projection false-target mode still blocked the strict claim. OP4 added a projection-frame lock. Under the locked OP4 measurement model, the synthetic observable-projection identification bridge passed the strict gate. OP5 repeated the projection-frame protocol on fresh holdout seeds. It reproduced a strong target-vs-null signal, but the strict false-positive gate was narrowly blocked by the permuted-projection control. OP6 added a frame-equivariance guard on fresh holdout seeds. Under the locked OP6 measurement model, the synthetic projection-frame / frame-equivariance guarded observable-projection identification bridge passed the strict gate, with all pairwise null margins passing and no false-target-nearest null exceeding the declared gate. The strongest admissible Stage 19 statement is therefore deliberately narrow: Stage 19 provides synthetic, protocol-bound support that a projection-frame and frame-equivariance guarded observable-projection identification bridge can preserve the finite native-6D orientation-sensitive carrier fingerprint against the declared OP6 null ladder. This is not an empirical observable validation. It is not a validation of physical spacetime, continuum quantum field theory, particle spectra, mass generation, Standard-Model recovery, BD17A, theta = 2/9, C3/A5 confirmation, or HLV ontology. The archive supports only a synthetic bridge-audit result under the stated finite computational protocols. Negative and blocked outcomes are intentionally preserved. OP1, OP2, OP3, and OP5 remain part of the audit record because they identify the residual failure modes that led to the stricter OP4 and OP6 guard designs. This archive is intended as reproducibility material for a later rigorous manuscript on the Stage 19 observable-projection audit.