Artian's Physical Terminal Reference Framework: A Carrier-Independent T-star/Tau Constructor Theorem for Activation-Gated Tests of QTT Axiom A1

When does a physical device contain two genuinely different clock terminals? This framework gives a carrier-independent constructor for a source terminal and a laboratory-written terminal. A four-history experiment becomes eligible only when a completed source event, continuous physical memory, reference-write graph, target-blind camera, detector degree, covariance, chronology, power, and ordinary-null model all pass their frozen gates. The clock statistic is \[ \boxed{ R_{\rm clk} =\frac{1}{C_KN_{\times}^{\rm ordinary}} \sqrt{\frac{|D_{\tau T}D_{T\tau}|} {|D_{\tau\tau}D_{TT}|}} } \] and the activated point fork is \[ \boxed{ H_{\rm ordinary}:R_{\rm clk}=1, \qquad H_{\rm QTT-A1}:R_{\rm clk}=\cos^q\!\left(\frac{\pi}{8}\right). } \] Version 1.1 adds the conventional-Hamiltonian completeness theorem: \[ \boxed{ G_{\rm PT}^{(1.1)} =G_{\rm PT}^{(1.0)}G_{\rm null}, \qquad G_{\rm null}=0 \Longrightarrow \mathrm{INELIGIBLE\_NO\_THEORY\_VERDICT}. } \] For rotating atom, spin, or ring-cavity platforms, the ordinary ledger begins from \[ H_{\rm rot}^{\rm std} =H_0-\Omega\cdot(L+F+\ell_F) +H_{\rm RR}+H_Z+H_{\rm Stark}+H_{\rm trap} +H_{\rm bs}+H_{\rm pol}+\cdots. \] Every applicable row must be included or bounded using pre-data parameter provenance and held-out controls. The machine-readable certificate now contains sixteen gates and rejects target-trained or post-result null construction. Concept DOI: 10.5281/zenodo.21739215 Main book: Quantum Traction Theory: Main Book v10.01 Website: quantumtraction.org Version 1.2: operation qualification and retained observations The numerical targets, ratio orientation, decision regions and historical seals are unchanged. A 22.5-degree or 45-degree component is no longer excluded by its angle alone: its physical transfer must be independently established. Common-operation cancellation and signal-preserving calibration are proved; both hypotheses must remain identifiable. Every acquired record is retained. After authorized unblinding, QTT-like, ordinary-like and unexpected outcomes are reported, including observations that do not qualify for a theory verdict. This is a prospective amendment, not a retrospective resealing of data. Clock, ruler and signal-recovery qualification. Both registered predictions are constructed at the raw detector with explicit proper-time, coordinate-time, worldline, ruler and synchronization conventions. Each relativistic effect is counted once. Independently calibrated corrections carry their uncertainties through the common comparison; the tested rows cannot train a normalization that forces either answer. Blind recovery is required separately for both alternatives at every predeclared operating setting. The inherited targets and decision rules remain fixed. The included design study makes sensitivity explicit: for the Two-Test Gram classifier, a total ratio uncertainty of 0.0013 gives approximately 88.9% QTT-branch and 94.1% ordinary-branch power under the stated ideal model. These are conditional detection probabilities, not probabilities that a theory is true, and each instrument retains its own classifier. The complete apparatus packet must be sealed before acquisition and target access; historical campaign rules remain preserved. Reconstruction entry point. Start with current/CURRENT_RELATIVISTIC_EXECUTION.md in the ZIP. The shared execution contract refers once to the development filename CURRENT_EXECUTION_RC.md; that reference means CURRENT_RELATIVISTIC_EXECUTION.md in this release. The packaged README names the correct file. This navigation correction does not change the deposited files, scientific requirements or historical seals.