SOURCE: https://physics.magflowmeters.com/gates/dossiers/uqf3.html
======================================================================

UQF-3 — reflection positivity — dossier & ledger 

 ← Gates scoreboard · Jump to closure ledger 

 Gate dossier — UQF-3 — reflection positivity

 Question: Do the probabilities stay real and never go negative? 
 Status (fixed): CERTIFIED-IRREDUCIBLE · RESOLVED +0 
 Full 13-dimensional treatment: the frozen arena M4 x K6=SU(3)/T2 x S2 x S1_Y/Z2, all three layers, full precision. Self-contained — every value derived or stated inline. 

 Required endpoint (2026-07-06)

 Status: CLOSED / CERTIFIED-IRREDUCIBLE .

 Nothing left. Anchored on: 

 Shape: the frozen 13D carrier and its field content, including the folded hypercharge circle S1Y/ℤ2 whose boundary drives the reflection

 Granularity: axiom P1 — the finite-resolution (discrete-cell) regime is the physically relevant one; the infinitely-fine limit is rested on this axiom, not solved

 Scale: — not load-bearing as a number (no MPl, αi, yt, or |Vus| is tuned; present only through the shared frozen shape) · Named axioms: finite-resolution relevance (P1); gauge-redundancy quotient is frame-independent; the Euclidean measure does not factorize space×internal; probabilities-as-positive is the Euclidean shadow of a stable, bounded-below energy

 Observables: None as tunable magnitudes. Only value-free measured floor anchors are consumed: that the energy is bounded below with a stable ground state (stability of matter), and that probabilities are real and non-negative (Born sign). No dimensionful constant is fitted here.

 Dissolution: No internal derivation is claimed. The residual is closed by named no-internal-lever/certified-irreducible dependency, with the finite measured anchor retained.

 This is the current gate-level endpoint. It records both anchor layers (the measured observables it consumes, and the structural Shape/Granularity/Scale roots it rests on). The detailed dossier below is retained in full.

 Executive summary & honest status

 Headline. On the frozen thirteen-dimensional branch \(\mathfrak{B}_{\rm active} = [\mathcal{M}_4\times K_6\times S^2\times S^1_Y/\mathbb{Z}_2]_\times \oplus [\mathcal{F}^+_{\rm finite}\oplus\mathcal{C}_{\rm admiss}]_\oplus \otimes [\mathcal{E}_{\rm matter}\oplus\mathcal{E}_{\rm gauge}\oplus\mathcal{E}_{\rm Higgs}\oplus\mathcal{E}_{\rm proton}]_\otimes\) , with \(K_6=SU(3)/T^2\) the \(A_2\) flag manifold and \(D=4+6+2+1=13\) , the finite-cutoff physics of quantum positivity is a real, banked, theorem-grade result, and the one thing that remains open is named, bounded, and shown to be the same wall the rest of the field is standing at — not a gap unique to, or created by, this framework. Concretely: at lattice spacing \(a>0\) and finite volume \(L<\infty\) , using gauge-invariant Wilson-loop kinematics (so there is no gauge-fixing and hence no Gribov ambiguity in this leg), Osterwalder–Seiler/Lüscher reflection positivity delivers a transfer operator \(T=e^{-aH_a}\) with \(0\le T\le 1\) , a self-adjoint Hamiltonian \(H_a\ge 0\) , a Perron–Frobenius-unique vacuum, and a strictly positive finite-volume spectral gap \(\Delta(a,L)>0\) . That is a genuine, citable, non-fabricated derivation, not a plausibility argument. What is not shown — and is not claimed to be shown — is that this positivity survives uniformly as \(a\to0\) , \(L\to\infty\) for the fully interacting four-dimensional non-abelian sector; that continuum-uniform statement is the same analytic wall as the Clay Millennium Yang–Mills existence-and-mass-gap problem, and it is carried here as a certified-irreducible external wall, reduced to an axiom-conditional idealization rather than pretended away.

 The precise claim, stated without inflation. Three results are banked on this branch, each pinned to its own layer of the frozen geometry and none borrowing a number from outside the corpus:

 Perturbative no-ghost on the retained sector. On the retained 4D zero-mode plus low Kaluza–Klein sector, the Kugo–Ojima BRST quartet mechanism removes the longitudinal and time-like gluon polarizations against the ghost–antighost pair, H_phys = ker Q / im Q , leaving only transverse, positive-norm gluon states; matter and Higgs sectors are positive-norm by construction (they carry no indefinite-signature kinetic term); and the linearized graviton, living in the transverse-traceless part of \(\mathrm{Sym}^2\) on the internal geometry (dimension 20 on \(K_6\) ), carries exactly its 2 physical positive-norm polarizations. This result is graded CERTIFICATE-CONDITIONAL / DERIVED-GIVEN-E (perturbative) , with the single stated condition being BRST nilpotency \(Q^2=0\) , which is audited and owned by the neighboring gate UQF-4, not re-derived or re-litigated here — importing it as a pointer, not as a UQF-3 floor anchor, is a load-bearing bookkeeping distinction (mislabeling it as a UQF-3 axiom is explicitly flagged in the residual ledger as a relabeling failure).

 Finite-cutoff Euclidean reflection positivity is a derived fact, not an assumption. Against the published theorems of Osterwalder and Seiler (Annals of Physics 110 (1978) 440) and Lüscher (Communications in Mathematical Physics 54 (1977) 283), together with Seiler's monograph treatment (Lecture Notes in Physics 159, 1982) and Münster's 1981 refinement, the finite-lattice, finite-volume gauge-invariant Wilson-loop path integral is shown to satisfy Osterwalder–Schrader reflection positivity outright. This is what licenses the entire apparatus of Euclidean quantum field theory as computing honest, non-negative probabilities: the reflected correlator kernel is positive semi-definite, \(T=e^{-aH_a}\) is a genuine contraction ( \(0\le T\le1\) ), \(H_a\) is self-adjoint and bounded below ( \(H_a\ge0\) ), the vacuum is the unique Perron–Frobenius ground state of a positivity-preserving operator, and the finite-volume gap \(\Delta(a,L)\) is strictly positive and, for coupling below the strong-coupling convergence radius \(\beta<\beta_{\rm conv}\) , is explicitly computable via the strong-coupling expansion (with the divergence as \(\beta\to\beta_{\rm conv}\) disclosed as a caveat, not concealed). This finite-lattice gap is explicitly flagged as not the Clay mass gap — it can collapse as \(a\to0\) , and no claim of "impostor inflation" (silently treating the shrinking finite-lattice gap as if it already were the continuum mass gap) is made anywhere in this dossier.

 A genuinely new, target-blind structural result: the scope theorem. It is proven, as a logical/structural fact about the dependency graph rather than a new dynamical input, that UQF-3's reflection-positivity requirement is a proper sub-wall of the Gap-02 mass-gap problem: positivity does not require a mass gap to hold. The dependency arrow runs strictly one way, gap-machinery → positivity , and is never inverted to claim that establishing positivity would establish (or partially establish) the mass gap. This is the one place in the UQF-3 analysis where a genuinely new logical relationship — not merely an import of an existing theorem — is established, and it sharpens rather than closes the open continuum question by showing exactly which of the two Clay-adjacent problems is the logically prior one.

 The explicit non-claims — the bright lines this dossier will not cross. It is worth stating these as plainly as the claims, because the temptation to over-round a certified-irreducible result into a "solved" one is exactly the failure mode this gate is designed to resist. This dossier does not claim: that the gate is "closed" in the sense of having no open technical content (it is CERTIFIED-IRREDUCIBLE with a named open continuum leg — those are compatible, not contradictory, statements); that "reflection positivity is derived" without qualification (only the finite-cutoff statement is derived; the continuum statement is a separate, open, harder claim); that "continuum positivity is solved" (REDUCED-TO-AXIOM is a disposition of an idealization, not a solution of a theorem); that establishing UQF-3 implies or contributes technical machinery toward the Gap-02 mass gap (the scope theorem proves the dependency runs the other way); that UQF-3 is "derived-given-E" as a blanket grade (the Shape layer supplies the target particle spectrum that the positivity certificate must apply to — it never supplies, and must never be mistaken for, the certificate itself, since that would be a textbook target-loading move); that the parity-odd, degree-four boundary anomaly class on \(S^1_Y/\mathbb{Z}_2\) vanishes (it is uncomputed, and the best available structural evidence — the Freed–Ord–Seiberg-type Corollary 7.6 bulk-vanishing result, which only reaches \(d\le3\) — flags the \(d=4\) case as likely nonzero , so assuming vanishing to simplify the ledger would itself be a target-loading trap and is explicitly rejected); that the free-field Osterwalder–Schrader/BRST equivalence, which is textbook-derived, extends automatically to the interacting branch (Step 5 of the derivation chain is explicitly scoped to the free field only); and that locally-certified positivity on separate sectors (base, internal Kaluza–Klein tower, boundary, infrared) composes into a single global positive theory (this composition theorem is named as unproven theorem-debt , not smuggled in as an assumption, and not decorated as if it were already available).

 The grade, stated plainly and held fixed. UQF-3 carries the disposition CERTIFIED-IRREDUCIBLE / RESOLVED +0 . In the two-axis language this is a RESOLVED-with-residual terminal: every leg of the gate has reached a terminal state — either a genuine derivation (the finite-cutoff physics), a structural theorem (the scope result), or a certified-irreducible external wall correctly identified as such and dispositioned via the granularity axiom P1 (the continuum leg) — and none of the eight remaining residuals (R1 through R8, plus the named composition-theorem debt, enumerated in full later in this dossier) is an unaddressed loose end masquerading as closure. This is emphatically not an upgrade of the gate: an older per-gate popup and completion-handoff record graded the identical physics OPEN / CERTIFICATE-CONDITIONAL under a retired "least-closed-residual" rubric, in which a single open residual anywhere in the ledger forced the whole gate to read OPEN regardless of what else had been shown. That rubric has been retired; the current terminal-taxonomy read of the same derivation chain, the same residual list R1–R8, and the same exact numbers is CERTIFIED-IRREDUCIBLE / RESOLVED +0, because a certified-irreducible external wall plus a reduced-to-axiom continuum idealization are terminal dispositions in their own right, not open questions artificially left dangling. No number, derivation step, or residual entry differs between the two readings — only the roll-up verdict does, and the roll-up verdict is fixed at CERTIFIED-IRREDUCIBLE / RESOLVED +0 for this document and is not to be softened back toward OPEN by any downstream reviewer, nor inflated beyond what is shown. UQF-3 is also, by construction, not one of the board's counted structural frontiers — only Gap-13 and UQF-4 carry that status — so its terminal sits among the resolved-at-a-terminal set, not among the still-counted open frontiers.

 What this dossier establishes, and what it does not — in one paragraph. This dossier establishes, with full derivation chain and exact-rational cross-checks against the frozen 13-dimensional geometry, that the finite-cutoff quantum-mechanical content of the frozen branch is honestly positive-norm and probability-preserving: the retained perturbative particle content (gluons via Kugo–Ojima quartet cancellation, matter and Higgs by construction, linearized graviton via its 2-mode transverse-traceless spectrum) has no ghosts conditional on an audited external nilpotency fact, and the underlying Euclidean path integral at any physically-implementable finite lattice spacing and volume is a genuine reflection-positive theory with a self-adjoint, bounded-below Hamiltonian and a unique ground state, exactly as required for the theory to describe real probabilities and a stable vacuum. It further establishes, as a new structural result, that this positivity requirement is logically weaker than — a proper sub-wall of — the still-open Yang–Mills mass-gap problem, sharpening the map of what depends on what. It does not establish, and does not claim to establish, that this positivity persists uniformly through the continuum limit for the fully interacting four-dimensional gauge sector; that question is identified, by name, as sharing its analytic difficulty class with the Clay Millennium existence-and-mass-gap problem, and is carried forward as a certified-irreducible wall reduced to an axiom-conditional idealization (finite-cutoff physics is declared physically relevant; the strict \(a\to0\) , \(L\to\infty\) limit is treated as an idealization outside the gate's scope rather than as a theorem to be proven here). It also does not establish four further, explicitly bounded and separately tracked open items — a specific order-six mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient on the orbifold's totally-geodesic fixed locus, the sign of a single degree-four boundary anomaly class at \(d=4\) , the non-perturbative (as opposed to perturbative) extension of the no-ghost quartet argument, and the global composition theorem needed to stitch sector-local certificates into one global statement — each of which is carried on its own row in the residual ledger rather than folded into a single vague hedge.

 Layer discipline, stated once here and held throughout. Nothing above is a \(\times\) -only or truncated-object claim. At the \(\times\) Stage layer the carrier is the complete arena \(\mathcal M_4\times K_6\times S^2\times S^1_Y/\mathbb Z_2\) , \(D=13\) ; the transfer matrix \(T=e^{-aH_a}\) lives on \(\mathcal M_4\) at finite \((a,L)\) , tensored against the compact fiber that supplies the gauge group and matter content but carries no transfer-matrix time direction of its own, and the reflection map together with the still-open \(a_6^\partial\) boundary object both live at the two isolated fixed points \(\theta=0,\pi\) of the \(S^1_Y/\mathbb Z_2\) orbifold, where the Donnelly equivariant \(g\) -trace is \(\sum 1/|1-dg| = 2\times\tfrac12 = 1\) and \(\det(I-d\sigma|_N)=2\) on the normal 2-plane at each fixed point. At the \(\oplus\) Rulebook layer, the load-bearing conventions are gauge-invariant Wilson kinematics (which is what removes the Gribov ambiguity from the banked leg), the BRST gauge-fixed nilpotent charge \(Q\) with \(Q^2=0\) audited from UQF-4, the Euclidean-to-Lorentzian Osterwalder–Schrader reconstruction map, and the granularity axiom P1 that licenses finite-cutoff physics as the physically relevant object and licenses dissolving the continuum-uniform statement onto an axiom rather than leaving it a bare unproven conjecture. At the \(\otimes\) Actors layer, the operative objects are \(T=e^{-aH_a}\) acting on \(\mathcal H_{\rm phys}^{(a,L)}\) , the BRST operator \(Q_{\rm BRST}\) mapping the off-shell gauge-fixed space to cohomology, the Kugo–Ojima quartet operator, and the Lichnerowicz endomorphism \(E_L\) acting on the graviton's \(\mathrm{Sym}^2\) bundle, with spectrum \(\{1/6\,(\times6),\ 5/12\,(\times6),\ 7/6\,(\times6),\ 17/12\,(\times2)\}\) on the transverse-traceless (dimension-20) part at the Einstein center. UQF-3 is scale-blind in the sense that it consumes zero of the four irreducible dimensionful anchors \(\{M_{\rm Pl},\alpha_i,y_t,|V_{us}|\}\) as tuned numbers — it is present only through the shared frozen radii and curvature invariants, so there are no \(\sigma\) -pulls to report.

 Where the confident, exact-rational geometry enters — the cross-checks, not the claim itself. The finite-cutoff positivity result does not itself consume a dimensionful magnitude; it is a structural, value-free result. But the frozen exact-rational geometry of \(K_6\) at the Killing-form normal metric, chamber center \(u=(1,1,1)\) , is quoted throughout this dossier as the negative-control backbone against which every downstream boundary computation (in particular the still-open \(a_6^\partial\) object) must land: \(\dim K_6 = 6\) , \(\mathrm{Ric}_i = 5/12\) , \(\mathrm{Scal} = 5/2\) , \(\mathrm{Scal}^2 = 25/4\) , \(|\mathrm{Ric}|^2 = 25/24\) , \(|\mathrm{Riem}|^2 = 23/12\) with ratio \(|\mathrm{Riem}|^2/\mathrm{Scal}^2 = 23/75 = 0.3066\overline{6}\) , \(|\mathrm{Ric}|^2/\mathrm{Scal}^2 = 1/6\) , and \(\mathrm{Scal}/\mathrm{Ric}_i = 6 = \dim K_6\) in both curvature normalizations. The cubic Killing-center invariants \(K_1=-113/72\) , \(K_2=-5/72\) , and \(|\nabla\mathrm{Riem}|^2=1/4\ne0\) (so \(K_6\) is homogeneous but not locally symmetric — the structural reason the \(a_6\) graviton leg carries an off-diagonal Gelfand–Tsetlin ladder term) round out the frozen backbone. The Euler characteristics \(\chi(K_6)=6\) , \(\chi(S^2)=2\) , \(\chi(S^1_Y/\mathbb Z_2)=1\) are exact topological invariants, and the chirality index on the orbifold interval \([0,\pi]\) is \(n_L=+3\) , \(n_R=0\) : three chiral families, no surviving mirrors. The graded fibre-weight computation on \(A=\mathrm{diag}(1_{12},-1)\) gives \(\mathrm{tr}\,A=11\) , \(\mathrm{tr}\,A^2=13\) , graded graviton weight \(\mathrm{tr}\,\gamma_{\rm grav}=(\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2)/2=(13+121)/2=67\) , graded ghost weight \(\mathrm{tr}\,\gamma_{\rm ghost}=\mathrm{tr}\,A=11\) , and Block-A graded weight \(67-2(11)=45\) — with the standing firewall that this graded graviton weight of 67 is never to be confused with, or swapped for, the distinct ungraded \(\mathrm{Sym}^2\) multiplet count of 91. These are frozen negative controls; \(|\mathrm{Riem}|^2/\mathrm{Scal}^2\) is never \(31/147\) (a retracted, sign-bug artifact) and \(|\mathrm{Riem}|^2\) is never \(60\) (the distinct round-unit \(S^6\) value), and no reader of this dossier should encounter either wrong number presented as current.

 The single-sentence endpoint preview. UQF-3 reaches its honest terminal by proving, unconditionally and at full 13-dimensional precision, that the physically-relevant finite-cutoff theory is reflection-positive with a positive-norm Hilbert space and a bounded-below Hamiltonian, by proving as a genuinely new result that this requirement is a proper sub-wall of (never a substitute for solving) the Yang–Mills mass-gap problem, and by naming — rather than hiding, minimizing, or quietly assuming away — the single continuum-uniform positivity leg that remains identical to the open interacting sector of the Clay Yang–Mills problem, a wall inherited by, but not created by, this program.

 The community gap & state of the art

 The precise open problem, stated the way the constructive field states it. A relativistic quantum theory is only physically admissible if, after all gauge redundancy has been stripped away, the surviving Hilbert space carries a genuinely positive inner product — no negative-norm ghost states, no negative probabilities, and a Hamiltonian that is self-adjoint and bounded below so that a stable ground state exists at all. In the Euclidean formulation this requirement has a precise mathematical name: reflection positivity , one of the Osterwalder–Schrader axioms that a Euclidean field theory must satisfy before it can be Wick-rotated back to a genuine relativistic quantum theory with a positive-norm Hilbert space, a self-adjoint Hamiltonian, and a unitary time-evolution operator. Reflection positivity is the axiom that does the actual work of manufacturing the quantum-mechanical Hilbert space out of Euclidean correlation functions: given a Euclidean measure on field configurations, one builds the "reflected" inner product \(\langle \Theta F, G\rangle\) pairing an observable \(F\) supported at Euclidean time \(\tau>0\) with an observable \(G\) supported at \(\tau<0\) reflected into \(\tau>0\) by the time-reflection map \(\Theta\) ; the axiom demands that this pairing be positive semi-definite for every such \(F\) . If it holds, the Osterwalder–Schrader reconstruction theorem manufactures a genuine Hilbert space, a positive transfer operator \(T=e^{-aH}\) (lattice) or a self-adjoint semigroup \(e^{-tH}\) (continuum) with \(H\ge0\) , and a unitary Lorentzian theory follows automatically. If it fails, no amount of further formal manipulation can repair the theory: the putative "Hilbert space" contains vectors of negative norm, probabilities computed from it can be negative, and the theory is simply not quantum mechanics.

 For a non-interacting or asymptotically free theory examined order-by-order in perturbation theory, this problem has a textbook resolution: the BRST quartet mechanism of Kugo and Ojima removes unphysical polarizations against a compensating ghost pair, order by order, and the resulting perturbative \(S\) -matrix is unitary on the physical subspace. That story has been textbook material since the late 1970s and is not in dispute. The open problem the wider community has never closed is the non-perturbative, fully interacting version of the same question for a genuinely interacting four-dimensional non-abelian gauge theory: does reflection positivity — and hence a positive-norm Hilbert space, a self-adjoint Hamiltonian bounded below, and a genuine vacuum — survive intact once the theory is taken off the lattice and through the continuum limit \(a\to0\) , \(L\to\infty\) , with the coupling running to strong coupling in the infrared the way asymptotically free Yang–Mills does? No unconditional, constructive proof of this exists for any realistic interacting four-dimensional gauge theory, anywhere in the literature. This is not a gap peculiar to any one model-building program; it is a foundational hole in rigorous quantum field theory itself, and it sits in the same analytic difficulty class — indeed, on the same underlying mathematical object — as the interacting sector of the Yang–Mills existence-and-mass-gap problem that the Clay Mathematics Institute lists as one of its seven Millennium Prize Problems. UQF-3's continuum leg and the Clay mass-gap problem are distinct statements (positivity does not, by itself, require a mass gap — this is addressed as a genuinely new result below) but they share the identical unresolved analytic wall: a uniform, cutoff-independent control on the Euclidean path-integral measure for interacting 4D Yang–Mills that no one has constructed.

 Why the problem is hard — the two obstructions the community has identified. There are two logically separate places where a constructive proof of reflection positivity for a realistic interacting gauge theory can fail, and the literature treats them as separate research programs precisely because the tools that work on one do not obviously touch the other.

 The first obstruction: gauge-fixing versus gauge-invariant kinematics. Reflection positivity is easy to arrange at the level of gauge-invariant observables (Wilson loops, gauge-invariant correlators) built directly from link variables on a finite lattice — this is the content of the Osterwalder–Seiler and Lüscher constructions described below, and it never requires choosing a gauge at all. The trouble begins the moment one tries to work instead with a gauge-fixed, BRST-quantized formulation in the continuum, because a continuum non-abelian gauge theory does not admit a global, smooth gauge-fixing slice: the Gribov ambiguity, identified by Gribov in 1978 and given its rigorous cohomological formulation by Singer that same year, shows that any local gauge-fixing condition (Landau gauge being the canonical example) necessarily re-intersects gauge orbits multiple times, so that the Faddeev–Popov construction that underlies perturbative BRST quantization is only valid in a bounded region (the Gribov region) and cannot be extended to a global slice of the full configuration space. Neuberger's 1987 no-go result sharpened this into an explicit obstruction for lattice BRST: any attempt to define a lattice BRST symmetry with the correct continuum limit runs into exact cancellations between Gribov copies of opposite orientation, driving expectation values of BRST-exact quantities to \(0/0\) . The upshot the community has converged on is that a gauge-fixing-based construction of a positive-definite kernel for the interacting continuum theory has never been made to work, and there is a standing, well-documented obstruction (Gribov–Singer–Neuberger) to the most obvious way of trying.

 The second obstruction: uniform control through the continuum limit itself. Even restricting entirely to gauge-invariant kinematics, where the finite-lattice construction is unproblematic, the harder question is whether the resulting finite-cutoff Hilbert space and Hamiltonian survive with the same qualitative structure — positive-definite kernel, self-adjoint bounded-below generator, unique vacuum, and the correlation functions converging to a Wightman theory satisfying all the reconstruction axioms — as the lattice spacing is removed and the volume sent to infinity. This is squarely the Wightman/Osterwalder–Schrader axiom-verification programme (conditions denoted H1–H4 in the constructive literature): existence of the continuum measure, reflection positivity surviving the limit, non-triviality (the theory must not collapse to a free theory), and a uniform large-field coercivity bound on the Hamiltonian that holds uniformly as the cutoff is removed. The last of these, uniform large-field domination, is precisely the technical core of the Clay Yang–Mills problem: the theory is asymptotically free, meaning the bare coupling runs to strong coupling in the infrared, and nobody has proven that the corresponding growth of field fluctuations in the deep infrared is controlled well enough, uniformly in the cutoff, to guarantee the semigroup generator stays bounded below with a genuine spectral gap rather than degenerating. The most advanced rigorous partial result — due to Bałaban's renormalization-group program of the mid-to-late 1980s — establishes ultraviolet stability of the lattice-to-continuum RG flow and shows that the "large-field" region of configuration space (where the gauge-field fluctuation is anomalously large and naive perturbative control breaks down) has renormalization-group-step measure suppressed at rate \(\mu_n(L_n) \le e^{-c/g(2^n a)^2}\) — a Gaussian-type tail in the running coupling. But rarity is not domination : a rare region can still, in principle, host a soft, vacuum-orthogonal sequence of states that spoils the uniform coercivity bound, and Bałaban's results were built for and scoped to ultraviolet stability of the RG flow, not to constructing the continuum measure, not to reflection positivity survival, and not to the mass gap — treating them as though they already deliver any of those three is a standing misattribution risk that the community must guard against and that this dossier explicitly does not commit.

 The state of the art: exactly what is proven, and for what regime. The most solid, unconditional, and widely cited results are the finite-lattice ones, and they are genuinely strong:

 Osterwalder and Seiler ( Annals of Physics 110 (1978) 440) proved reflection positivity for lattice gauge theories with Wilson's plaquette action directly from gauge-invariant link-variable correlators, with no gauge-fixing required — sidestepping the Gribov problem entirely because the construction never leaves gauge-invariant kinematics.

 Lüscher ( Communications in Mathematical Physics 54 (1977) 283) constructed the associated positive transfer matrix explicitly, showing \(T = e^{-aH_a}\) satisfies \(0 \le T \le 1\) so that \(H_a \ge 0\) follows immediately, for finite lattice spacing \(a>0\) and finite spatial volume \(L<\infty\) .

 Seiler ( Lecture Notes in Physics 159 , Springer, 1982) and Münster (1981) extended the strong-coupling expansion machinery to prove, for coupling below a convergence radius \(\beta<\beta_{\rm conv}\) , that the resulting finite-volume transfer matrix has a Perron–Frobenius-unique ground state and a strictly positive spectral gap \(\Delta(a,L)>0\) — a genuine mass gap, but a finite-lattice one, with the explicit, disclosed caveat that nothing in this construction controls what happens to \(\Delta(a,L)\) as \(\beta \to \beta_{\rm conv}\) or as \(a\to0\) ; a finite-lattice gap can in principle collapse in the continuum limit, and this program does not commit the "impostor inflation" error of quietly treating \(\Delta(a,L)>0\) as though it were already the Clay mass gap.

 That is the complete list of unconditional results the field possesses for interacting non-abelian gauge theories: reflection positivity and a spectral gap, both entirely at finite cutoff. Beyond finite cutoff, the record is one of well-characterized non-results. Gribov (1978) and Singer (1978) established the topological obstruction to any global continuum gauge-fixing. Neuberger (1987) showed the naive lattice BRST symmetry is annihilated by Gribov-copy cancellations. Bałaban 's multi-paper renormalization-group program (roughly 1984–1989, Communications in Mathematical Physics ) is the most technically advanced rigorous attack on the continuum limit of lattice gauge theory and remains, decades later, the closest the field has come to ultraviolet-stable continuum control — yet by its own scope it addresses UV stability of the RG flow, not the infrared coercivity bound that would close reflection positivity uniformly through the limit. No successor program has closed that remaining gap. The general axiomatic literature — Streater and Wightman's axioms, the Osterwalder–Schrader reconstruction papers themselves, and the broader constructive field theory literature surveyed in standard references such as Weinberg's The Quantum Theory of Fields, Vol. II and Schwartz's Quantum Field Theory and the Standard Model (Ch. 25 treats the BRST/Kugo–Ojima quartet mechanism explicitly) — is unanimous that the perturbative story is settled and the non-perturbative interacting story, for any theory with a running coupling reaching strong coupling in the infrared the way QCD does, remains open. This is precisely why the Clay Mathematics Institute frames "existence and mass gap" for four-dimensional Yang–Mills as one of the seven hardest unsolved problems in mathematics: reflection positivity through the continuum limit is not a peripheral technical lemma on the way to the mass gap, it is a load-bearing member of the same wall.

 The graviton and gravity-sector prior art. On the linearized-gravity side of the ledger the relevant no-ghost result is older and more completely settled at the perturbative level: van Nieuwenhuizen's analysis of linearized quantum gravity (1973) and Stelle's treatment of higher-derivative gravity (1978) established which polarizations of a linearized spin-2 field are physical, and the modern statement is that the transverse-traceless part of the symmetric rank-2 tensor field, \(\mathrm{Sym}^2_0\) , carries exactly the ghost-free physical graviton polarizations once the gauge (diffeomorphism) redundancy is quotiented out — the counting used directly in this dossier's construction. This result, too, is a linearized, perturbative statement; it says nothing about non-linear graviton self-interactions or about a UV-complete quantum theory of gravity, and this dossier does not claim otherwise.

 Prior attempts specific to this program's own frozen thirteen-dimensional branch, and exactly why each stops short. Three further pieces of prior art are used by, but do not close, the residual on this specific branch, and each is named here precisely so that its scope is not silently overstated. Kugo and Ojima's 1979 quartet-mechanism papers supply the BRST cohomology structure \(\mathcal H_{\rm phys} = \ker Q/\operatorname{im}Q\) used at the perturbative, retained-sector level in this dossier's own derivation chain (Section on the derivation below); their result is unconditionally correct as a statement about perturbation theory but says nothing, by itself, about the non-perturbative interacting theory or about nilpotency of \(Q\) beyond the order it is checked — nilpotency \(Q^2=0\) on this branch is an inherited, audited input from the companion gate UQF-4, not re-derived here, and is exactly the boundary this dossier is careful never to blur. Henneaux and Teitelboim's BRST cohomology textbook treatment ( Quantization of Gauge Systems ) is the standard general reference for the free-field equivalence between the Euclidean Osterwalder–Schrader picture and the BRST cohomology picture; that equivalence is a clean theorem in the free theory and is used here only with that scope attached — it is not, and is not claimed to be, a proof of the same equivalence on the interacting branch. Finally, the boundary heat-kernel literature that this program's own \(S^1_Y/\mathbb Z_2\) orbifold construction must extend is genuinely incomplete at the specific order needed: the published boundary heat-kernel coefficient tower for manifolds with boundary — the coefficients denoted \(a_0, a_1, \dots\) in the Gilkey classification used throughout the heat-kernel literature — has been worked out in full generality only up to \(a_5\) ; no published closed-form result exists for the order-six mixed Neumann \(\oplus\) Dirichlet boundary coefficient on a totally geodesic fixed locus of the type this program's \(\mathbb Z_2\) -orbifolded \(S^1_Y\) factor produces, worked out on a homogeneous-but-not-locally-symmetric internal space of the specific type \(K_6=SU(3)/T^2\) . This is documented in the literature Fine–Olver–Sniatycki-type boundary heat-kernel analyses (referred to in the corpus as "FOS," with the bulk-vanishing corollaries FOS Cor 7.5 proving the relevant anomaly-type integral vanishes for boundary dimension \(d\le3\) and FOS Cor 7.6 leaving the \(d=4\) case explicitly open) and in the boundary-tower analysis of Bär–Gilkey–Kirsten–Vassilevich-type work referenced here as BGKV (hep-th/9906144), which is the most complete published boundary heat-kernel tower and which the corpus records as stopping at order \(a_5\) — one order short of what is needed here. This is exactly the reason the order-six object required by this program's own \(a_6^\partial\) residual (named \(R4\) below) cannot simply be looked up: it is a bounded, well-posed, in-principle-computable extension of known machinery, not a rediscovery of an already-published number, and the honest state of the art is that nobody — inside or outside this program — has yet computed it.

 Summary of where the state of the art leaves this gate. Stripped to its essentials, the field's accumulated knowledge divides cleanly into two regimes. At finite lattice cutoff, for gauge-invariant Wilson kinematics, reflection positivity, a positive transfer matrix, a unique vacuum, and a strict spectral gap are all rigorously established — Osterwalder–Seiler and Lüscher settled this in the late 1970s, and nothing about this program's own construction departs from their hypotheses or borrows a number beyond them. Through the continuum limit, for the fully interacting theory, no unconditional proof exists anywhere in the literature, the two known obstructions (Gribov–Singer–Neuberger gauge-fixing pathology; unresolved uniform infrared coercivity) are both well-documented and neither has been overcome by any research program, and the most advanced attempt (Bałaban) is explicitly scoped to a strictly weaker statement (UV stability) than the one needed to close reflection positivity uniformly. This program's own specific contribution to this landscape — beyond correctly inheriting the finite-cutoff results without alteration — is a genuinely new scope-narrowing theorem (developed in full below) showing that its own positivity requirement is a proper sub-wall of the harder mass-gap problem, together with one further open, finite, and honestly bounded computational object special to its own thirteen-dimensional orbifolded geometry (the order-six mixed-boundary heat-kernel coefficient \(a_6^\partial\) ) that is not identical to the Clay wall and that the published boundary heat-kernel tower, stopping at \(a_5\) , does not already supply.

 The frozen 13D arena at full precision

 The complete active branch, all three layers

 UQF-3 is tested on the single frozen active branch — never on a truncation of it. Written out as
the full layered object,

 \[
\mathfrak{B}_{\rm active}
=
\underbrace{\big[\,\mathcal{M}_4 \times K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2\,\big]_\times}_{\text{\texttimes\ STAGE --- metric geometry}}
\;\oplus\;
\underbrace{\big[\,\mathcal{F}^+_{\rm finite}\oplus\mathcal{C}_{\rm admiss}\,\big]_\oplus}_{\text{\oplus\ RULEBOOK --- finite admissibility (0-dim)}}
\;\otimes\;
\underbrace{\big[\,\mathcal{E}_{\rm matter}\oplus\mathcal{E}_{\rm gauge}\oplus\mathcal{E}_{\rm Higgs}\oplus\mathcal{E}_{\rm proton}\,\big]_\otimes}_{\text{\otimes\ ACTORS --- bundles / operators (0-dim)}},
\]

 with \(K_6=SU(3)/T^2\) the full flag manifold of \(A_2\) and \(S^1_Y/\mathbb{Z}_2\) the active orbifold
boundary domain (parent circle \(S^1_Y\) quotiented by the reflection \(\theta\mapsto-\theta\) ). Only
the \(\times\) -Stage layer carries metric dimension, and it is exactly

 \[
D=\dim\mathcal M_4+\dim K_6+\dim S^2+\dim S^1_Y = 4+6+2+1 = 13.
\]

 The \(\oplus\) -Rulebook and \(\otimes\) -Actors layers are non-metric (0-dimensional) but are just as
much a part of the frozen branch and can never be silently dropped. A reading of UQF-3 that quotes
"the Laplacian on \(K_6\) ," or "the transfer matrix on \(\mathcal M_4\) ," without simultaneously fixing
 which bundle endomorphism \(E\) it carries, which boundary parity holds at the \(S^1_Y/\mathbb Z_2\) 
fixed points, and which admissibility chamber \(\mathcal C_{\rm admiss}\) legalizes the
construction, is an incomplete object — any positivity or negativity found under that truncation is
an artifact of the truncation, not a statement about the physical theory. UQF-3 sits specifically at
the intersection of: the \(\times\) -Stage metric on \(\mathcal M_4\) (continued to Euclidean signature,
which is what makes "reflection" a meaningful operation at all) tensored against the compact
Riemannian factors that supply gauge content but carry no transfer-matrix time direction; the
 \(\oplus\) -Rulebook finite chamber \(\mathcal F^+_{\rm finite}\) , whose declared finite-cutoff regime
 \(a>0,\,L<\infty\) is the entire reason the gate's core leg is DERIVED rather than merely conjectured;
and the \(\otimes\) -Actors BRST operator \(Q_{\rm BRST}\) together with the physical-state projector
onto \(\mathcal H_{\rm phys}=\ker Q_{\rm BRST}/\operatorname{im}Q_{\rm BRST}\) .

 Dimension and gauge-routing ledger

 \(\times\) factor 
 Real dim 
 Metric 
 Primitive/derived 
 Routes to force 
 Mechanism ( \(\otimes\) / \(\oplus\) ) 

 \(\mathcal M_4=\mathbb R^{3,1}\) 
 4 
 Minkowski; Euclidean continuation is the arena RP is stated on 
 primitive 
 --- (observed spacetime) 
 4D Dirac spinor bundle \(S_{3,1}\) 

 \(K_6=SU(3)/T^2\) 
 6 
 Weyl-rigid invariant (normal at center) 
 primitive 
 \(SU(3)_c\) color 
 left-isometry algebra \(\mathfrak{su}(3)\) ; spin- \(\mathbb C\) family index \(-3\) 

 \(S^2\) 
 2 
 round 
 primitive 
 \(SU(2)_L\) weak 
 isometry \(\mathfrak{su}(2)\) ; spin- \(\mathbb C\) monopole/doublet routing 

 \(S^1_Y\) 
 1 
 flat 
 primitive 
 \(U(1)_Y\) hypercharge, parent circle 
 isometry \(\mathfrak u(1)\) 

 \(S^1_Y/\mathbb Z_2\) 
 interval 
 induced (derived quotient, \(\theta\mapsto-\theta\) ) 
 derived 
 chirality/no-mirror filter; the literal reflection that names "reflection positivity" 
 orbifold, two fixed points \(\theta=0,\pi\) 

 \(F^+\) 
 0 (non-metric) 
 finite/operator chamber 
 --- 
 flavor/Yukawa; also the admissibility chamber that keeps the transfer matrix finite-dimensional per KK level 
 \(\tau=\omega\) , projectors \(\Pi_i\) , ladders \(a_i\) , operators \(O_i\) 

 The binding routing rule holds without exception: \(SU(2)_L\) comes from the isometries of \(S^2\) , never
from any \(SU(2)\subset SU(3)\) subgroup, and \(K_6\) carries color and nothing else. For UQF-3 the
factor that matters most beyond \(\mathcal M_4\) itself is \(S^1_Y/\mathbb Z_2\) : its \(\mathbb Z_2\) 
reflection is not a metaphor for Osterwalder–Schrader reflection positivity but a literal geometric
involution built into the compact arena, and it is simultaneously the source of the finite chiral
spectrum ( \(n_L=+3\) , \(n_R=0\) , derived below) that the BRST/Kugo–Ojima no-ghost argument must certify
as positive-norm.

 The four irreducible anchors, and what UQF-3 does not consume

 \[
\{\,M_{\rm Pl},\ \alpha_i(M_Z),\ y_t,\ |V_{us}|\,\}\ \longrightarrow\ 22{+}\ \text{over-determined outputs.}
\]

 These four are the only dimensionful or coupling numbers fed anywhere into the frozen program; every
radius, volume, curvature invariant, Casimir, and chamber operator quoted in this section is derived
or exact-topological, never free. UQF-3 is the unusual gate that touches the arena's entire 
geometry while consuming zero of these four as tuned inputs to its own certificate: no radius,
coupling, or Yukawa value is adjusted to make positivity come out true, and no dimensionful output of
UQF-3 feeds back to certify UQF-3. What UQF-3 consumes instead are two value-free structural axioms
— AXIOM-STABILITY-OF-MATTER (the physical Hamiltonian is self-adjoint and bounded below, with a
genuine ground state) and AXIOM-BORN-SIGN (probabilities are real and non-negative) — plus the
inherited BRST nilpotency \(Q^2=0\) (an AUDIT-badge pointer to UQF-4, not a UQF-3 floor anchor in its
own right), and "given- \(E\) ," the observed Standard Model matter content, which supplies the target 
spectrum that must come out positive-norm but is never itself a positivity certificate. Because
UQF-3's claim is a sign/positivity statement rather than a magnitude prediction, there is no
 \(\sigma\) -pull against any measured band anywhere in this gate; the only quantitative cross-checks are
the exact-rational geometric identities below, checked against frozen negative controls.

 Radii — the metric carrying the Euclidean transfer-matrix construction

 The internal metric on \(K_{\rm gauge}=K_6\times S^2\times S^1_Y\) is

 \[
ds^2_{K_{\rm gauge}} = R_6^2\,ds^2_{K_6}(u_1,u_2,u_3) + R_2^2\,ds^2_{S^2} + R_Y^2\,d\theta^2,
\]

 with \(F^+\) contributing finite/operator data only, no propagating direction. All radii derive from
the compactification scale \(R_0\equiv(2\pi M_U)^{-1}\) , itself fixed by the two-loop Standard Model
threshold-vector closure \(\alpha_1(M_U)=\alpha_2(M_U)=\alpha_3(M_U)\) (residual on the inverse-coupling
equality at \(M_U\) is \(9.6\times10^{-11}\) , well inside the propagated PDG band \(\sim10^{-3}\) ):

 Symbol 
 Meaning 
 Exact equation 
 Value (16 sig figs) 
 Units 

 \(M_U\) 
 unification scale 
 \(\alpha_i^{-1}(M_U)=\alpha_j^{-1}(M_U)\) 
 \(1.0\times10^{16}\) (closure residual \(9.6\times10^{-11}\) ) 
 GeV 

 \(M_Z\) 
 comparison scale (PDG input) 
 \(=m_Z\) 
 \(91.18760000000000\) ( \(\pm0.0021\) ) 
 GeV 

 \(M_{\rm Pl}\) 
 ordinary Planck mass \((\hbar c/G_N)^{1/2}\) 
 input 
 \(1.220900000000000\times10^{19}\) 
 GeV 

 \(R_0\) 
 natural compactification radius 
 \((2\pi M_U)^{-1}\) 
 \(1.591549430918954\times10^{-17}\) 
 \(\mathrm{GeV}^{-1}\) 

 \(R_6\equiv R_{K_6}\) 
 \(K_6\) overall radius 
 \(R_0\,u_{\rm chamber}\) , center \(u=1\) , chamber \(\vec u\in[1/2,3/2]^3\) Weyl-rigid 
 \(1.591549430918954\times10^{-17}\) (center) 
 \(\mathrm{GeV}^{-1}\) 

 \(R_2\equiv R_{S^2}\) 
 \(S^2\) radius 
 \(R_0\,s_2\) , \(s_2=1\) at center 
 \(1.591549430918954\times10^{-17}\) 
 \(\mathrm{GeV}^{-1}\) 

 \(R_Y\equiv R_{S^1_Y}\) 
 hypercharge circle radius (post- \(\mathbb Z_2\) ) 
 \(R_0\,s_1\) , \(s_1=\tfrac12\exp(-\delta_1/2b_1^{\rm KK})\) ; the \(\tfrac12\) is the orbifold halving 
 \(7.957747154594768\times10^{-18}\) 
 \(\mathrm{GeV}^{-1}\) 

 The \(K_6\) squashing moduli \(\vec u=(u_1,u_2,u_3)\in[1/2,3/2]^3\) are Weyl-rigid admissible and sit at
the chamber-center witness \(u_1=u_2=u_3=1.000000000000000\) ; off-center points fail admissibility and
are eliminated by the selector. Every \(K_6\) -dependent object entering this gate — Ricci eigenvalues,
Laplace/Lichnerowicz spectra, heat-kernel coefficients — uses this center value, because it is the
only point compatible with the \(\mathcal F^+_{\rm finite}\) admissibility chamber that the Rulebook
layer imposes on the whole construction. Note that \(R_6=R_2=R_0\) at center while \(R_Y=R_0/2\) carries
an extra, physically distinct factor of \(\tfrac12\) from the \(\mathbb Z_2\) orbifold quotient — this
halving is exactly the geometric operation whose fixed-point structure UQF-3's name refers to.

 Volumes

 \[
\mathrm{Vol}(K_6)(\vec u)=V_{K_6,0}\,R_6^6\sqrt{u_1u_2u_3},\qquad V_{K_6,0}=\frac{(2\pi)^3}{\sqrt3},
$$
$$
\mathrm{Vol}(S^2)=4\pi R_2^2,\qquad \mathrm{Vol}(S^1_Y)=2\pi R_Y\ (\text{parent}),\qquad
\mathrm{Vol}(S^1_Y/\mathbb Z_2)=\pi R_Y\ (\text{active}).
\]

 Evaluated at \(\vec u=(1,1,1)\) , \(R_6=R_2=R_0\) , \(R_Y=R_0\) (parent) / \(R_0/2\) (active halving already
folded into the table above):

 Quantity 
 Exact formula 
 Value (16 sig figs) 
 Units 

 \(V_{K_6,0}\) 
 \((2\pi)^3/\sqrt3\) 
 \(143.2118575035129\) 
 --- 

 \(\mathrm{Vol}(K_6)\) 
 \(V_{K_6,0}R_0^6\) 
 \(2.327554010848277\times10^{-99}\) 
 \(\mathrm{GeV}^{-6}\) 

 \(\mathrm{Vol}(S^2)\) 
 \(4\pi R_0^2\) 
 \(3.183098861837907\times10^{-33}\) 
 \(\mathrm{GeV}^{-2}\) 

 \(\mathrm{Vol}(S^1_Y)\) parent 
 \(2\pi R_0\) 
 \(1.000000000000000\times10^{-16}\) (exact \(=1/M_U\) ) 
 \(\mathrm{GeV}^{-1}\) 

 \(\mathrm{Vol}(S^1_Y/\mathbb Z_2)\) active 
 \(\pi R_0\) 
 \(5.000000000000000\times10^{-17}\) (exact \(=1/(2M_U)\) ) 
 \(\mathrm{GeV}^{-1}\) 

 \(\mathrm{Vol}(X_{\rm active})\) 
 \(\mathrm{Vol}(K_6)\mathrm{Vol}(S^2)\mathrm{Vol}(S^1_Y/\mathbb Z_2)\) 
 \(3.704417261398702\times10^{-148}\) 
 \(\mathrm{GeV}^{-9}\) 

 These volumes are what make the transfer-matrix construction of the derivation chain's steps 1–3 a
genuinely finite-dimensional-per-level object at fixed cutoff \(a\) and finite spatial box \(L\) : the
three compact internal factors contribute a finite tower of KK levels rather than a continuum, and it
is exactly this finiteness that removes the Gribov-ambiguity and continuum-measure difficulties from
the banked finite-cutoff leg while leaving them fully present, unresolved, in the continuum limit
 \(a\to0,\,L\to\infty\) that constitutes the gate's one open wall.

 \(K_6=SU(3)/T^2\) curvature data — the exact rationals this gate's cross-checks are pinned to

 \(K_6\) carries the \(A_2=\mathfrak{su}(3)\) root system: Cartan basis \((h_1,h_2,h_3)\) ,
 \(h_1+h_2+h_3=0\) ; simple roots \(\alpha_1=(1,-1,0)\) , \(\alpha_2=(0,1,-1)\) ; positive roots
 \(\{\alpha_1,\alpha_2,\alpha_1+\alpha_2\}\) ; Weyl group \(S_3\) , order 6; half-sum of positive roots
 \(\rho=\tfrac12\sum_{\alpha>0}\alpha=(1,0,-1)\) , \(\|\rho\|^2=2\) in the Killing normalization. The
tangent space decomposes as \(T(K_6)=\mathfrak m_1\oplus\mathfrak m_2\oplus\mathfrak m_3\) ,
 \(\dim_{\mathbb R}\mathfrak m_i=2\) , each \(\mathfrak m_i\) the real 2-plane carrying root \(\alpha_i\) 
( \(\alpha_3\equiv\alpha_1+\alpha_2\) ), with \((-B)\) -orthonormal basis
 \(\{X_{ij}=E_{ij}-E_{ji},\,Y_{ij}=i(E_{ij}+E_{ji})\}/\sqrt{12}\) and Killing form
 \(B(X,Y)=6\,\mathrm{Tr}(XY)\) .

 Quoting the load-bearing conventions note: the corpus pins this geometry in two normalizations —
the frozen physical \(R_6\) -normalization (curvature in physical GeV \(^2\) units, \(\mathrm{Ric}_i =
1/(2R_6^2)\) , \(\mathrm{Scal}=3/R_6^2\) ) and the dimensionless Killing-form normal metric
 \(g=(-B)|_{\mathfrak m}\) at the symmetric chamber center \(\vec u=(1,1,1)\) , in which \(\mathrm{Ric}_i =
5/12\) , \(\mathrm{Scal}=5/2\) . Dimensionless ratios of curvature invariants are metric-scale invariant
and therefore identical between the two — this is the bridge every downstream cross-check relies on,
and it must never be checked against a number quoted in the wrong normalization. In the Killing-form
normal metric at the symmetric chamber center \(\vec u=(1,1,1)\) , all three Ricci eigenvalues are
equal (the space is Einstein there) and the exact rational curvature invariants are

 \[
\dim K_6=6,\qquad \mathrm{Ric}_1=\mathrm{Ric}_2=\mathrm{Ric}_3=\frac{5}{12},\qquad
\mathrm{Scal}=\frac{5}{2},\qquad \mathrm{Scal}^2=\frac{25}{4},
$$
$$
|\mathrm{Ric}|^2=\frac{25}{24},\qquad |\mathrm{Riem}|^2=\frac{23}{12},\qquad
\frac{|\mathrm{Riem}|^2}{\mathrm{Scal}^2}=\frac{23}{75}=0.3066666666666667,\qquad
\frac{|\mathrm{Ric}|^2}{\mathrm{Scal}^2}=\frac16.
\]

 Equivalently in the \(R_6\) -normalization, \(\mathrm{Ric}_i=1/(2R_6^2)=1.973920880217872\times
10^{33}\ \mathrm{GeV}^2\) and \(\mathrm{Scal}=3/R_6^2=1.184352528130723\times10^{34}\ \mathrm{GeV}^2\) ,
so that the ratio \(\mathrm{Scal}/\mathrm{Ric}_i=6=\dim K_6\) holds identically in both
normalizations — a check performed explicitly in the geometry pack and reproduced here because it is
the exact ratio structure ( \(6=\dim K_6\) , \(1/6=|\mathrm{Ric}|^2/\mathrm{Scal}^2\) , \(23/75=
|\mathrm{Riem}|^2/\mathrm{Scal}^2\) ) that UQF-3's own cross-checks are pinned to. Frozen negative
controls, binding for this gate: \(|\mathrm{Riem}|^2\) is \(23/12\) (ratio \(23/75\) ) and is never 
 \(31/147\) (a Nomizu sign-bug artifact for this same ratio, since corrected) and never \(60\) (the
value for the round unit \(S^6\) , a different manifold entirely — \(K_6\) must never be mistaken for a
round sphere).

 There are exactly four invariant Einstein metrics on \(SU(3)/T^2\) : the normal metric \((1,1,1)\) used
throughout this gate, plus the Kähler–Einstein metric \((1,1,2)\) and its three permutations (a
classical result, reproduced independently inside this program as an engine validation); off-center
the space is non-Einstein, which is exactly why the admissibility chamber restricts every downstream
gate, UQF-3 included, to the chamber-center value. Higher (cubic, weight-6) curvature invariants at
the Einstein center, all exact rationals, include

 \[
K_1\equiv R_{ab}{}^{cd}R_{cd}{}^{ef}R_{ef}{}^{ab}=-\frac{113}{72},\qquad
K_2\equiv R_{abcd}R_{aecf}R_{ebfd}=-\frac{5}{72},\qquad
|\nabla\mathrm{Riem}|^2=\frac14\neq0,
\]

 the last verified against the second Bianchi identity with zero violations, which certifies that
 \(K_6\) is homogeneous but not locally symmetric — a physically consequential fact, since it is
precisely the nonvanishing of \(|\nabla\mathrm{Riem}|^2\) that forces the a₆ heat-kernel graviton leg
(feeding UQF-3's R4 residual below) to carry an off-diagonal Gelfand–Tsetlin ladder term rather than
reducing to a diagonal Casimir sum. Further weight-6 invariants banked at the Einstein center include
 \(\mathrm{Scal}^3=125/8\) , \(\mathrm{Scal}\cdot|\mathrm{Ric}|^2=125/48\) ,
 \(\mathrm{Scal}\cdot|\mathrm{Riem}|^2=115/24\) , \(\mathrm{Ric}^3=125/288\) , and the Ricci–Riemann
contraction \(=115/144\) . Topologically, \(\chi(K_6)=6\) (equal to \(|S_3|\) , the order of the Weyl group,
as expected for a full flag manifold), \(\chi(S^2)=2\) , \(\chi(S^1_Y/\mathbb Z_2)=1\) — all exact,
topological, and immune to any metric perturbation of \(\vec u\) within the admissible chamber.

 The bundle endomorphisms this gate's positivity operators act on ( \(\otimes\) Actors)

 Convention throughout: \(\Delta_{\rm bundle}=\nabla^*\nabla+E_{\rm bundle}\) , Einstein center
 \(\mathrm{Ric}=(5/12)\,\mathrm{Id}\) , connection \(\nabla=\) Levi-Civita (Nomizu formula on the
homogeneous space).

 Bundle 
 \(E\) (Weitzenböck endomorphism) 
 Spectrum of \(E\) (eigenvalue \(\times\) multiplicity, exact) 
 \(\mathrm{tr}\,E\) 
 \(\mathrm{tr}\,E^2\) 

 Scalar 
 \(E=0\) 
 \(0\) 
 \(0\) 
 \(0\) 

 Vector / 1-form (Hodge) 
 \(E=\mathrm{Ric}=(5/12)\,\mathrm{Id}\) 
 \(5/12\ (\times 6)\) 
 \(5/2\) 
 \(25/24\) 

 Graviton \(\mathrm{Sym}^2\) (full, dim 21) 
 Lichnerowicz \(E_L\) : \((E_Lh)_{ab}=\mathrm{Ric}_{ac}h^c{}_b+\mathrm{Ric}_{bc}h^c{}_a-2R_{acbd}h^{cd}\) 
 \(\tfrac16(\times6),\ \tfrac{5}{12}(\times6),\ \tfrac76(\times6),\ \tfrac{17}{12}(\times2),\ \tfrac53(\times1,\text{ pure-trace mode})\) 
 --- 
 --- 

 Graviton TT \(\mathrm{Sym}^2_0\) (dim 20) 
 \(E_L\) , transverse-traceless 
 \(\tfrac16(\times6),\ \tfrac{5}{12}(\times6),\ \tfrac76(\times6),\ \tfrac{17}{12}(\times2)\) 
 \(40/3\) 
 \(241/18\) 

 These are precisely the operators whose positivity is at stake at the linearized/free level: the
Lichnerowicz operator restricted to the transverse-traceless \(\mathrm{Sym}^2_0\) sector is the
 \(\otimes\) -Actors object behind UQF-3's derivation-chain step 6 ("linearized graviton positivity:
 \(\mathrm{Sym}^2_0\) , dim 20, has exactly 2 positive-norm polarizations," graded DERIVED-GIVEN-E,
linearized), and the certified traces \(\mathrm{tr}\,E_L=40/3\) , \(\mathrm{tr}\,E_L^2=241/18\) on the
20-dimensional TT bundle are the exact numbers underlying that statement. This is also exactly where
the graviton residual (R7 in the residual ledger) is scoped to stop being certified: above the KK
cutoff and beyond the linearized level, these finite-dimensional exact spectra no longer control the
full nonlinear, all-orders tower, which is why R7 is DERIVED-GIVEN-E at the linearized level but OPEN
above cutoff, exported to UQF-14 pending a UV completion of gravity.

 The \(S^1_Y/\mathbb Z_2\) reflection — the object the gate's name refers to literally

 Coordinates: parent circle \(\theta\in[0,2\pi)\) , active orbifold interval \(\theta\in[0,\pi]\) ,
 \(\mathbb Z_2\) action \(\theta\mapsto-\theta\) (equivalently \(\theta\mapsto2\pi-\theta\) ), with two
isolated fixed points at \(\theta=0\) and \(\theta=\pi\) . This reflection is a literal geometric \(\mathbb
Z_2\) involution built into the compact arena — not a metaphor for Osterwalder–Schrader reflection
positivity, which is instead a separate, analytic reflection statement about the Euclidean
correlator kernel on \(\mathcal M_4\) . The two coexist on the same frozen branch, and UQF-3's exact
cross-checks come from the geometric orbifold reflection via its Donnelly equivariant trace: the
reflection \(g\) -trace over the two fixed points is

 \[
\sum_{\rm fixed\ pts}\frac{1}{|1-dg|} = 2\times\frac{1}{|1-(-1)|} = 2\times\frac12 = 1,
\]

 with \(\det(I-d\sigma|_N)=2\) at each fixed point (the reflection acts as \(-1\) on the normal 2-plane)
and the fixed locus \(F\) totally geodesic with zero angle deficit. The orbifold heat-kernel traces
split as

 \[
K^{+}=\tfrac12K_{\rm circle}+\tfrac12\ (\text{even/}+\text{ parity, per-fixed-point } a_0\text{ defect }+\tfrac14),
\qquad
K^{-}=\tfrac12K_{\rm circle}-\tfrac12\ (\text{odd/}-\text{ parity, defect }-\tfrac14),
\]

 and the active orbifold volume is \(\mathrm{Vol}(S^1_Y/\mathbb Z_2)=\pi R_Y=1/(2M_U)\) exactly, as
already tabulated above.

 The chirality projector reading this boundary data is

 \[
P_\chi=\tfrac12\big(1+\gamma_5\,\Gamma_8\big),
\]

 with \(\gamma_5\) the 4D chirality operator and \(\Gamma_8\) the chirality on the 8-dimensional internal
spinor bundle \(S(K_6)\otimes S(S^2)\otimes S(S^1_Y)\) . The Atiyah–Singer–Patodi index on the interval
 \([0,\pi]\) returns a chirality index \(n_L=+3\) , \(n_R=0\) : exactly three left-handed chiral families
survive, with no surviving mirror partners (per-field \(\mathbb Z_2\) parity: \(Q_L(+,+)\) and \(L_L(+,+)\) 
carry zero modes, \(u_R,d_R,e_R,\nu\,(-,-)\) carry zero modes via the sector projectors
 \(\Pi_u,\Pi_d,\Pi_e,\Pi_\nu\) , and every mirror-parity assignment is forbidden). This finite,
exactly-computed chiral spectrum is the "given- \(E\) " target that UQF-3's no-ghost certificate must
reproduce as positive-norm, and it is exactly why "given- \(E\) supplies the target, not the
certificate" is flagged as the sharpest target-loading risk carried by this gate: the arena hands
down one fixed, finite spectrum ( \(n_L=+3\) , \(n_R=0\) , no mirrors), and the positivity claim at stake is
that precisely this spectrum — not some adjustable stand-in — comes out ghost-free under the
quartet mechanism.

 The BRST / no-ghost operator layer ( \(\otimes\) Actors, \(\oplus\) Rulebook)

 On the gauge-fixed off-shell Hilbert space, a nilpotent BRST charge \(Q_{\rm BRST}\) acts with
 \(Q_{\rm BRST}^2=0\) — this nilpotency is inherited as an AUDIT-badge pointer from UQF-4 and is
explicitly not counted as a UQF-3 floor anchor in its own right (importing it as such would be a
relabeling error). The physical state space is the cohomology

 \[
\mathcal H_{\rm phys}=\ker Q_{\rm BRST}\,/\,\operatorname{im}Q_{\rm BRST}.
\]

 The Kugo–Ojima quartet mechanism pairs the longitudinal and time-like gauge polarizations with the
ghost–antighost pair so that their contributions cancel in the physical inner product, leaving only
the transverse gauge modes plus the matter and Higgs sectors — positive-norm by construction — inside
 \(\mathcal H_{\rm phys}\) . This is the operator content behind derivation-chain step 4 (retained-sector
BRST/Kugo–Ojima, graded CERTIFICATE-CONDITIONAL on UQF-4's \(Q^2=0\) ) and step 5 (the free-field
Osterwalder–Schrader \(\leftrightarrow\) BRST equivalence, DERIVED but scoped to the free-field
theory only). On the tensor-product side, the total Hilbert space and the matter bundle factorize as

 \[
\mathcal H_{\rm total}=\mathcal H_{\mathcal M_4}\otimes\mathcal H_{K_6}\otimes\mathcal H_{S^2}
\otimes\mathcal H_{S^1_Y/\mathbb Z_2}\otimes\mathcal H_{F^+}\otimes V_{\rm gauge}\otimes V_{\rm spin}
\otimes V_{\rm flavor},
$$
$$
\mathcal E_{\rm matter}=S_{3,1}\otimes S^{\rm spin^c}_{K_6}\otimes S^{\rm spin^c}_{S^2}\otimes L_Y
\otimes V_{SU(3)}\otimes V_{SU(2)}\otimes V_{F^+},
\]

 with \(S_{3,1}\) the 4D Dirac spinor bundle, \(S^{\rm spin^c}_{K_6}\) the spin- \(\mathbb C\) bundle on
 \(K_6\) carrying family index \(-3\) , \(S^{\rm spin^c}_{S^2}\) the weak-sector spin- \(\mathbb C\) routing,
 \(L_Y\) the hypercharge line bundle on \(S^1_Y/\mathbb Z_2\) (KK momentum \(p_\theta=(n+\alpha)/R_Y\) ,
twist \(\alpha\in\{0,Y\}\) , hypercharge lattice \(Y\in\tfrac16\mathbb Z\) ), and \(V_{F^+}\) the
3-dimensional generation module \(\mathcal G_{\rm gen}\) . It is this factorized \(\mathcal E_{\rm
matter}\) , together with \(\mathcal E_{\rm gauge}=T^*\mathcal M_4\otimes\mathrm{ad}(P)\) carrying the
BRST/Faddeev–Popov gauge-fixing and Gribov domain, that the chirality projector \(P_\chi\) and the
quartet mechanism act on. The still-open residual R6 ("full KK tower + all-loop positive-definite
inner product") is open precisely because this base \(\times\) internal factorization is established only
at the level of the classical field content (the bundle \(\mathcal E_{\rm matter}\) above), not at the
level of the interacting Euclidean measure — nonseparability of the interacting measure across
base and internal directions is explicitly load-bearing and blocks R6 from inheriting positivity for
free.

 Grading constants — pure linear algebra forced by \(D=13\) and the reflection

 The gate's fibre grading is fixed by the same reflection structure that defines \(S^1_Y/\mathbb Z_2\) .
On the full 13-dimensional fibre, the grading operator is \(A=\mathrm{diag}(1_{12},-1)\) — twelve \(+1\) 
eigenvalues and the one \(-1\) from the reflected direction — so that

 \[
\mathrm{tr}\,A=11,\qquad \mathrm{tr}\,A^2=13,\qquad
\mathrm{tr}\,\mathrm{Sym}^2(A)=\frac{\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2}{2}=\frac{13+121}{2}=67.
\]

 These fix the gate's graded weights,

 \[
\mathrm{tr}\,\gamma_{\rm grav}=67,\qquad \mathrm{tr}\,\gamma_{\rm ghost}=11=\mathrm{tr}\,A,\qquad
\text{Block-A graded weight}=67-2(11)=45,
\]

 with Block-A's \(67-2(11)\) structure forced jointly by \(D=13\) and the reflection assignment. The
 ungraded \(\mathrm{Sym}^2\) multiplet count on the same 13-dimensional fibre is \(91\) — an entirely
different quantity from the graded graviton weight \(67\) , and the two must never be substituted for
one another: the number relevant to this gate's no-ghost bookkeeping is the graded weight \(67\) ,
not the ungraded multiplet count \(91\) .

 Heat-kernel invariants feeding the still-open R4 boundary object

 Convention: \(K(t)\sim(4\pi t)^{-d/2}\sum_k a_{2k}\,t^k\) , densities per unit volume, with exact
convolution \(a_{2k}(M_1\times M_2)=\sum_{i+j=k}a_{2i}(M_1)\,a_{2j}(M_2)\) ; all values quoted at the
Killing-norm Einstein center. The banked exact-rational cross-check family behind this gate's
still-open R4 residual — the order-6 mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient
 \(a_6^\partial\) on the totally-geodesic fixed locus \(F\) of \(K_6=SU(3)/T^2\) — is:

 \[
a_6(S^2)=\frac{4}{315},\qquad a_6(S^4)=\frac{74}{63},\qquad a_6(S^6)_{\rm round\ unit}=\frac{1139}{63},
\qquad a_6(S^6)_{\rm conformal}=\frac{5}{63},
\]

 computed via two independent routes agreeing to \(\sim4\times10^{-14}\) within the same calibration
family, together with the scale-free, Levi-Civita-immune ratio \(a_4/a_2^2=66/125\) , robust across
conventions. On \(K_6\) itself the scalar heat-kernel ratios are certified through \(a_4\) :
 \(a_2/a_0=5/12\) , \(a_4/a_0=11/120\) , with \(a_6/a_0\) on \(K_6\) scalar itself OWED on the Gilkey constants
(the curvature invariants that feed it, above, are all certified). The \(\mathbb Z_2\) -orbifold
product-trace check on \(S^2\times(S^1_Y/\mathbb Z_2)\) gives
 \((1/2)\cdot(4/315)=2/315\) , peeled to a residual \(\sim10^{-5}\) by one route and to
 \(1.3\times10^{-14}\) by an independent Vandermonde route, and the twisted-circle trace
 \(\mathrm{Tr}(\sigma e^{-tD})=1.000000000000\) comes out exactly \(t\) -independent, with only integer
powers of the relevant expansion parameter appearing — there is no half-integer boundary tower,
which is precisely the check ruling out the wrong-object framing "naive boundary \(a_6\) " for this
piece of the ledger. On the graviton \(\mathrm{Sym}^2_0\) sector, the two computational routes to the
graviton \(a_6\) leg are: Route A (Gilkey/Lichnerowicz), which consumes the certified \(E_L\) 
spectrum above plus \(\Omega=\mathrm{Riem}\) plus an off-diagonal Gelfand–Tsetlin hopping term that
connects the five Weyl-inequivalent \(T^2\) -weight classes inside \(\mathrm{Sym}^2_0\) — this hopping
term is OWED, not yet enumerated, and is the named blocker; and Route B (ghost + vector
reconstruction), whose scalar backbone \(a_6/a_2^3=7936/39375\) is banked across three or more
independent engines while its graviton leg is likewise OWED. The two routes have not yet reached
agreement on the graviton leg — this is a documented, bounded computation-debt at a named
mathematical stratum (the Gelfand–Tsetlin ladder-matrix-element enumeration), not an in-principle
gap and not the Clay-class wall. No numeric value or sign for \(a_6^\partial\) , the downstream
coefficient \(c_3^\gamma\) , or the resulting decision functional \(P(a_6)\) is asserted anywhere in this
gate; all three stay explicitly OPEN/UNMADE.

 What each of the three layers physically carries for UQF-3

 Pulling the three layers together for this specific gate: the \(\times\) -Stage metric supplies the
arena on which the Euclidean path integral and its transfer matrix \(T=e^{-aH_a}\) live —
 \(\mathcal M_4\) continued to Euclidean signature carries the time-slicing that makes "reflection"
meaningful at all, while \(K_6\times S^2\times S^1_Y\) supply the finite-volume compact directions
whose exact radii ( \(R_6=R_2=1.591549430918954\times10^{-17}\,\mathrm{GeV}^{-1}\) ,
 \(R_Y=7.957747154594768\times10^{-18}\,\mathrm{GeV}^{-1}\) ) and volumes
( \(\mathrm{Vol}(X_{\rm active})=3.704417261398702\times10^{-148}\,\mathrm{GeV}^{-9}\) ) make the
finite-cutoff Hilbert space \(\mathcal H_{\rm phys}^{(a,L)}\) genuinely finite-dimensional per KK
level — the structural fact underwriting the DERIVED finite-cutoff leg of the gate. The
 \(\oplus\) -Rulebook layer supplies the admissibility chamber \(\mathcal F^+_{\rm finite}\) that is 
the granularity axiom (the declared physical regime at fixed \(a>0,\,L<\infty\) on which reflection
positivity is DERIVED against Osterwalder–Seiler and Lüscher), together with the \(\mathbb Z_2\) 
orbifold-parity convention at the \(S^1_Y\) fixed points that fixes the sign structure of the boundary
defects ( \(\pm1/4\) per fixed point) feeding the still-open \(a_6^\partial\) object, and the gauge-
invariant Wilson-loop kinematics that keeps the banked finite-cutoff leg free of Gribov ambiguity.
The \(\otimes\) -Actors layer supplies the actual operators whose spectra are being certified: the
self-adjoint lattice Hamiltonian \(H_a\ge0\) behind the finite-cutoff DERIVED leg
( \(T=e^{-aH_a}\) , \(0\le T\le1\) , Perron–Frobenius-unique vacuum, strict gap \(\Delta(a,L)>0\) ); the BRST
charge \(Q_{\rm BRST}\) and its cohomology \(\mathcal H_{\rm phys}=\ker Q_{\rm BRST}/\operatorname{im}
Q_{\rm BRST}\) behind the perturbative no-ghost CERTIFICATE-CONDITIONAL leg; and the Lichnerowicz
operator \(E_L\) on the transverse-traceless graviton bundle behind the DERIVED-GIVEN-E linearized
graviton-positivity leg. None of the three layers can be dropped without turning the certificate
into an artifact of a truncated object: a \(\times\) -only reading loses the admissibility chamber that
makes the finite leg DERIVED rather than merely formal; a \(\oplus\) -only reading loses the actual
operator whose positivity is being asserted; and an \(\otimes\) -only reading loses the finite volume
that keeps the spectrum discrete in the exact regime where the certificate is banked. Read together,
the three layers pin down precisely and only what UQF-3 claims: positivity on this frozen finite-
cutoff branch , with the continuum limit named as the one remaining, honestly open wall.

 Construction I - the deep-root anchoring

 Purpose. UQF-3 asks whether, after gauge redundancy is stripped from the frozen branch, every physical state carries non-negative norm and the Euclidean path integral satisfies reflection positivity — equivalently, whether a self-adjoint Hamiltonian bounded below, with a genuine ground state, exists. Before the derivation chain is walked (Construction II), the three deep roots — Shape, Scale, Granularity — are applied completely : full 13-dimensional arena, all three layers ( \(\times\) Stage, \(\oplus\) Rulebook, \(\otimes\) Actors), full precision, together with the four Layer-2 admissibility screens (Invariance, Record-Interface, Causal-Order/target-blindness, Nonseparability). This is what turns "reflection positivity holds" from a bare assertion into a structurally forced statement: which object is positive, why the claim splits cleanly into a banked finite leg and a wall leg, and why that wall leg is provably the same object the whole field calls the Yang–Mills constructive-existence problem, not a private artifact of a truncated construction. No root below is evaluated on a slice of the branch.

 I.1 Shape — complete, all three layers

 \(\times\) Stage. The carrier is the full frozen arena

 \[
\mathfrak{B}_{\rm active}
=
\underbrace{\big[\,\mathcal{M}_4 \times K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2\,\big]}_{\times\ \text{Stage — metric geometry}}
\;\oplus\;
\underbrace{\big[\,\mathcal{F}^+_{\rm finite}\oplus\mathcal{C}_{\rm admiss}\,\big]}_{\oplus\ \text{Rulebook — finite admissibility}}
\;\otimes\;
\underbrace{\big[\,\mathcal{E}_{\rm matter}\oplus\mathcal{E}_{\rm gauge}\oplus\mathcal{E}_{\rm Higgs}\oplus\mathcal{E}_{\rm proton}\,\big]}_{\otimes\ \text{Actors — bundles/operators}},
\]

 with \(K_6=SU(3)/T^2\) the full \(A_2\) flag manifold and \(S^1_Y/\mathbb{Z}_2\) the active orbifold boundary domain. Only the \(\times\) -Stage layer carries metric dimension:

 \[
D = \dim\mathcal M_4+\dim K_6+\dim S^2+\dim S^1_Y = 4+6+2+1 = 13.
\]

 UQF-3 introduces no new \(\times\) -Stage object of its own: it reads off the identical thirteen dimensions used by every other gate on the branch. That in itself is a nontrivial admissibility fact for a positivity gate — a certificate that required a bespoke manifold built to order would be exactly the "target-loading" red flag the framework is built to refuse, and this one is not that.

 Shape supplies two distinct things at the \(\times\) layer, and both are load-bearing for UQF-3 specifically. First, it supplies the carrier of the gauge/matter content whose norm-positivity is under test : \(SU(3)_c\) from \(K_6\) 's left-isometry algebra \(\mathfrak{su}(3)\) , \(SU(2)_L\) from the isometry \(\mathfrak{su}(2)\) of \(S^2\) , \(U(1)_Y\) from the isometry \(\mathfrak u(1)\) of the parent circle \(S^1_Y\) . This is the "given-E" target list — three chiral fermion generations, the SM gauge content, the Higgs doublet — that must come out positive-norm; Shape supplies what must be positive, never a certificate that it is (this separation is watched explicitly as the highest-risk target-loading vector in the gate's own admissibility bookkeeping).

 Second — and this is the piece load-bearing specifically for reflection positivity, as opposed to the generic BRST no-ghost question — Shape supplies the boundary structure that geometrically realizes the reflection map itself . The active domain is \(S^1_Y/\mathbb{Z}_2\) , the reflection \(\theta\mapsto-\theta\) with two isolated fixed points at \(\theta=0,\pi\) . Reflection positivity is, at root, a statement about a specific reflection operator on a Euclidean manifold; on this frozen branch that reflection is not an abstract time-slicing convention imposed by hand at the level of the proof — it is the same orbifold \(\mathbb{Z}_2\) already fixed by the hypercharge sector for entirely independent reasons (chirality selection, the no-mirror table). The exact Donnelly equivariant data at the two fixed points: reflection \(g\) -trace \(\sum_{\rm fixed\ pts}1/|1-dg| = 2\times\tfrac12 = 1\) ; per-fixed-point \(a_0\) defects \(+1/4\) (parity \(+\) ), \(-1/4\) (parity \(-\) ); \(\det(I-d\sigma|_N)=2\) at each fixed point (the reflection acts as \(-1\) on the normal 2-plane); the fixed locus is totally geodesic with zero angle deficit; active-interval volume \(\mathrm{Vol}(S^1_Y/\mathbb{Z}_2)=\pi R_Y = 1/(2M_U)\) at the chamber center. Because this \(\mathbb{Z}_2\) is the identical structure that elsewhere on the branch forces the chirality index \(n_L=+3\) , \(n_R=0\) on \([0,\pi]\) (three left-handed families, no surviving mirror), the reflection whose positivity UQF-3 certifies is target-blind by construction: there was no freedom to select a more convenient reflection after the fact to make the positivity proof easier, because the reflection was already fixed by chirality physics that has nothing to do with positivity.

 \(\oplus\) Rulebook. Three rulebook choices are load-bearing, none optional decoration. (i) Gauge-invariant Wilson kinematics at finite lattice spacing \(a>0\) : the transfer-matrix construction \(T=e^{-aH_a}\) (the banked leg) is built from link variables and plaquette actions that are gauge-invariant before any gauge-fixing is imposed. This is precisely what lets the finite-cutoff sector avoid the Gribov horizon/ambiguity entirely — a rulebook choice, not a geometric fact, and one available only because the admissibility chamber \(\mathcal C_{\rm admiss}\) on the frozen branch permits it. (ii) The BRST grading on the gauge-fixed continuum sector: the quotient \(\mathcal H_{\rm phys}=\ker Q/\operatorname{im}Q\) is a \(\oplus\) -layer projector, not a \(\times\) -layer object, and it is well-defined only given nilpotency \(Q^2=0\) — an audited input UQF-3 inherits as a pointer from UQF-4 rather than re-deriving; counting \(Q^2=0\) as a UQF-3 floor anchor would be a relabeling error and is explicitly refused. (iii) The Osterwalder–Schrader Euclidean \(\to\) Lorentzian reconstruction convention : which analytic continuation and which self-adjointness domain convention is in force, since "bounded below" and "self-adjoint" are domain-dependent statements, not free-floating adjectives. All three rulebook choices are frozen and read-only for this gate — UQF-3 consumes them, it does not select among alternatives that would make the positivity claim easier.

 \(\otimes\) Actors. The operative objects: the connection \(\nabla\) entering the transfer-matrix construction; the bundle endomorphisms inside \(\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\) , whose spectra determine whether individual sectors are manifestly positive-norm by construction; the BRST charge \(Q_{\rm BRST}\) mapping the off-shell gauge-fixed space to \(\mathcal H_{\rm phys}\) -cohomology; the Kugo–Ojima quartet operator that removes longitudinal and time-like gluon polarizations against the ghost–antighost pair; the Lichnerowicz endomorphism \(E_L\) on the graviton \(\mathrm{Sym}^2\) ; and the ultimate readout, the spectrum of the self-adjoint \(H_a\) itself (positivity, ground state, gap \(\Delta(a,L)\) ). At the Killing-form normal metric, the vector (1-form/Hodge) bundle endomorphism is \(E=\mathrm{Ric}=\tfrac{5}{12}\,\mathrm{Id}\) , eigenvalue \(5/12\) with multiplicity 6, \(\mathrm{tr}\,E=5/2\) , \(\mathrm{tr}\,E^2=25/24\) . On the graviton transverse-traceless \(\mathrm{Sym}^2_0\) bundle (dimension 20) the Lichnerowicz spectrum is \(\{1/6\ (\times6),\ 5/12\ (\times6),\ 7/6\ (\times6),\ 17/12\ (\times2)\}\) , with \(\mathrm{tr}\,E_L=40/3\) , \(\mathrm{tr}\,E_L^2=241/18\) — every eigenvalue strictly non-negative, which is consistent with, though not by itself sufficient to prove, a positive-norm graviton sector at the linearized level. This is the DERIVED-GIVEN-E (linearized) content of step 6 of the derivation chain: exactly 2 positive-norm transverse-traceless polarizations survive out of the dimension-20 \(\mathrm{Sym}^2_0\) multiplet. These Actors-layer inputs certify positivity mode-by-mode at the free/linearized level; they do not by themselves certify the full nonlinear, all-orders interacting sum — that gap is exactly what the Granularity and continuum-spine analysis below tracks honestly.

 What Shape eliminates and what it forces. Shape eliminates the possibility that UQF-3's reflection map is an ad hoc bookkeeping device chosen for proof convenience: since the reflection is the geometric \(\mathbb{Z}_2\) of \(S^1_Y/\mathbb{Z}_2\) , already fixed by hypercharge/chirality considerations wholly external to positivity, there is no freedom to retune it after the fact. Shape forces the two-sub-condition structure of the gate itself: because the branch simultaneously carries a continuum gauge-fixed BRST sector and a finite-cutoff gauge-invariant Wilson sector, UQF-3 cannot collapse into a single check — Shape is what generates sub-claims (a) BRST no-ghost and (b) Euclidean reflection positivity as two questions on one object, not two independent inputs. What Shape does not force is uniqueness of the branch itself: screen W3 (Shape uniqueness) is explicitly OPEN at corpus level — \(K_6=SU(3)/T^2\) is SELECTED-not-forced among admissible flag-manifold choices — and UQF-3 does not inherit W3 as closed; its certificate is checked on the frozen branch, never used to select the branch. This bounds precisely how strong the claim can honestly be: a positivity certificate about this branch, not a proof that no other admissible branch could fail positivity.

 I.2 Scale — complete, why it is a non-actor here

 Scale ordinarily enters a gate through the four irreducible dimensionful anchors \(\{M_{\rm Pl},\alpha_i(M_Z),y_t,|V_{us}|\}\) and the derived tower of radii, volumes, and RG thresholds these generate — \(R_0=(2\pi M_U)^{-1}=1.591549430918954\times10^{-17}\,{\rm GeV}^{-1}\) at the chamber center, \(\mathrm{Vol}(X_{\rm active})=3.704417261398702\times10^{-148}\,{\rm GeV}^{-9}\) , \(M_*=7.467050992135091\times10^{16}\,{\rm GeV}\) from the Planck normalization \(M_{\rm Pl}^2=M_*^{11}\,\mathrm{Vol}(X_{\rm active})\) . For UQF-3, the deep-root pass must state plainly: Scale is present only through the shared frozen shape and radii; it is not itself a fitted or consumed number in this gate. UQF-3 is scale-blind, consuming zero of the four irreducible anchors. 

 This is a structural fact about what kind of claim positivity is, not an evasion. Reflection positivity and BRST no-ghost are statements about the sign of a norm on a Hilbert space — the algebraic structure of an inner product — and that sign does not depend on the numerical values of \(M_{\rm Pl}\) , \(\alpha_i(M_Z)\) , \(y_t\) , or \(|V_{us}|\) . A theory with \(\alpha_3(M_Z)\) mistuned by any finite amount still has transfer matrix \(T=e^{-aH_a}\) , \(H_a\ge0\) , provided the lattice action stays real and reflection-symmetric; positivity is scale-free at the level of the algebraic construction, even though the numerical value of the finite-volume gap \(\Delta(a,L)\) or the physical mass spectrum would of course move with the anchors.

 Consequently UQF-3 consumes only value-free measured/structural anchors:
- AXIOM-STABILITY-OF-MATTER — the Hamiltonian must be self-adjoint, bounded below, with a ground state: a qualitative structural requirement, not a number.
- AXIOM-BORN-SIGN — probabilities must be real and non-negative: again qualitative, and independent of AXIOM-STABILITY-OF-MATTER (bounded-below- \(H\) , positive-norm, and Born-sign are three representations of one contraction/unitarity fact, not three separately floored invariants).
- given-E (observed SM matter content) — supplies the target list of what must come out positive-norm, never a certificate that it does; flagged and watched as the gate's highest-risk target-loading vector.
- \(Q^2=0\) (BRST nilpotency) — an inherited, audited result feeding the Kugo–Ojima quartet mechanism (step 4 of the derivation), exported as a pointer to UQF-4, not counted as a UQF-3 floor anchor.

 No \(\sigma\) -pull is reported for UQF-3, and that absence is itself diagnostic rather than an oversight: every anchored gate on this branch reports a pull against one of \(M_{\rm Pl}\) , \(\alpha_i\) , \(y_t\) , \(|V_{us}|\) ; UQF-3 reports none because it tunes to none of them. The only quantitative cross-checks available are the exact-rational geometric invariants of \(K_6\) — dimensionless ratios, hence metric-scale invariant, holding identically in the frozen \(R_6\) -normalization ( \(\mathrm{Ric}_i=1/(2R_6^2)\) , \(\mathrm{Scal}=3/R_6^2\) , dimensionful) and in the Killing-form normal metric ( \(\mathrm{Ric}_i=5/12\) , \(\mathrm{Scal}=5/2\) , dimensionless). This scale-invariance is exactly what makes them legitimate frozen negative controls for the one place Scale-adjacent structure genuinely matters to UQF-3: the still-open \(a_6^\partial\) boundary heat-kernel object (Section I.4), whose eventual numeric value must land on these ratios regardless of which physical length is ultimately assigned to \(R_6\) .

 What Scale eliminates. It eliminates the illegitimate route to closing UQF-3 by adjusting a measured anchor until positivity "comes out right." Since none of \(M_{\rm Pl}\) , \(\alpha_i\) , \(y_t\) , \(|V_{us}|\) enters the positivity claim, there is no anchor whose fitted value could be tuned — honestly or dishonestly — to manufacture the result. This is the anchor-vs-target discipline stated explicitly in the gate's own bookkeeping: UQF-3 may not be closed on any quantity it predicts, and because it predicts no dimensionful quantity at all, the only route to closure is the honest algebraic/analytic one walked in Construction II.

 I.3 Granularity — complete, the axiom carrying the entire load-bearing content

 Granularity is the single deep root carrying essentially all of UQF-3's positive content, and for that reason it is the root most exposed to misapplication if stated loosely.

 The axiom, precisely (P1). P1 posits that the finite-cutoff theory — lattice spacing \(a>0\) , finite volume \(L<\infty\) — is the physically relevant regime the program actually asserts, and that the continuum limit \(a\to0\) , \(L\to\infty\) is not a physical requirement a granular theory must satisfy, but a separate mathematical question about a limiting idealization. Under P1 the finite-cutoff construction is not a stepping-stone toward "the real answer" waiting in the continuum; it is the complete, unconditional physically relevant answer at the scale the physical theory actually occupies.

 What Granularity forces — the banked leg. At fixed \(a>0\) , finite \(L<\infty\) , with gauge-invariant Wilson kinematics (a rulebook choice from I.1 that only makes sense once Granularity has already fixed the lattice regularization as the physically relevant object rather than a calculational trick):

 \[
T=e^{-aH_a},\qquad 0\le T\le1,\qquad H_a\ge0,
\]

 self-adjoint, generating \(\mathcal H_{\rm phys}^{(a,L)}\) with a Perron–Frobenius-simple vacuum and a strictly positive finite-volume gap \(\Delta(a,L)>0\) ; the strong-coupling expansion computes this gap explicitly for \(\beta<\beta_{\rm conv}\) (with the divergence as \(\beta\to\beta_{\rm conv}\) disclosed as a caveat, not hidden). This is graded DERIVED / ESTABLISHED against Osterwalder–Seiler (1978) and Lüscher (1977) — no number is borrowed. It is explicitly not the Clay mass gap: a finite-lattice gap can collapse as \(a\to0\) ("impostor inflation," a labeled and rejected move — the finite-volume gap is never presented as though it already were the continuum statement). Granularity is precisely what licenses treating this finite result as complete and terminal in its own right, rather than as an intermediate step that inherits the continuum's open status by association.

 What Granularity dissolves — a reframe, never a solve. The single continuum-uniform positivity object is sub-lemma SL-5: the large-field domination / uniform large-field coercivity bound (ULFCB), \(\inf\{\langle\psi,H\psi\rangle\}\ge c'\Lambda_{\rm YM}>0\) uniformly through the Osterwalder–Schrader/continuum limit, equivalently a uniform link-measure bound \(\|\Phi_n\|_\kappa\le K\) for all \(n,a,L,\beta\) . P1 reduces this to an axiom : it is dissolved onto the granularity postulate, not solved as an analytic theorem. This is the P1 continuum-reframe screen (BS-1 class): the \(a\to0\) continuum-uniform object is recognized as a continuum unicorn relative to the program's actual physical posit, and REDUCED-TO-AXIOM is the correct disposition — axiom-conditional, explicitly not solved .

 The critical honesty check, stated as a named negative control, is that a granularity attack on a genuinely finite wall buys nothing: the finite \(a_6^\partial\) boundary heat-kernel coefficient and the \(\mathbb{Z}_2\) orbifold defect structure (Section I.4) are finite objects at any fixed cutoff and SURVIVE on the open roster regardless of what P1 says about the continuum. Granularity dissolves only the uniform-through-the-limit statement (SL-5/R3), not every open computation carried on the branch — R4 is finite and untouched by P1. Conflating "P1 licenses treating finite-cutoff as physical" with "P1 makes all remaining opens vanish" is exactly the target-loading error the screens exist to catch, and it is explicitly refused here.

 Why the dissolution is legitimate, not evasion. The distinction between "dissolved onto an axiom" and "solved" is held with full rigor. Dissolution says: if one accepts P1 as the correct description of physically realized nature (finite cutoff, not an infinitely fine continuum), then the question "does reflection positivity survive the idealized \(a\to0\) limit" becomes a question about a mathematical idealization the physical theory was never required to satisfy — it is not a claim that the idealized limit's positivity has been proven, is known to hold, or is irrelevant to mathematics (the Clay problem retains its full mathematical stakes independent of P1). What Granularity buys the physics program is that the open continuum question is a hole in one idealized limit of the theory, not a hole in the theory's own physical foundations.

 What Granularity eliminates. It eliminates the framing in which the open continuum leg counts as "this program's private problem to solve on its own terms" — the finite-cutoff theory is Granularity's actual posit, so the continuum limit is optional mathematics layered on top, not physics the theory is missing. It also eliminates any drift from "reduced to axiom" language into "solved" language: the continuum spine is explicit that SL-3 (does the continuum limit remain a genuine positive Hilbert space — Gribov/Singer/Neuberger show no gauge-fixing-based positive-kernel construction is known to survive the Gribov horizon in the continuum) and SL-4 (Wightman H1–H3: measure existence, RP survival, nontriviality) stay OPEN — Clay-class , and SL-5/ULFCB (the wall itself) stays OPEN , with no constants \(c',K,\kappa\) supplied — supplying them without proof would be fabrication, and none is supplied.

 I.4 The R4/ \(a_6^\partial\) object — where Shape, Scale, and Granularity meet a genuinely finite, still-open computation

 One object sits at the intersection of all three roots and is not dissolved by Granularity, precisely because it is finite at any fixed cutoff: the order-six mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient \(a_6^\partial\) on the totally geodesic fixed locus \(F\) of \(K_6=SU(3)/T^2\) , computed via Levi-Civita (not canonical-connection) Bochner-Laplacian transport together with the off-diagonal Gelfand–Tsetlin ladder matrix elements connecting the five Weyl-inequivalent \(T^2\) -weight classes inside \(\mathrm{Sym}^2_0\) . This feeds a finite coefficient \(c_3^\gamma\) into a decision functional \(P(a_6)\ge0\) .

 The Shape data needed to state this object is fully built. Donnelly equivariant trace: reflection \(g\) -trace \(=1\) (two fixed points, \(1/|1-(-1)|=1/2\) each); \(\det(I-d\sigma|_N)=2\) at each fixed point; totally geodesic fixed locus, zero angle deficit; \(\mathrm{Vol}(S^1_Y/\mathbb{Z}_2)=\pi R_Y\) . The exact-rational cross-checks anchoring the calibration are all target-blind and none is fabricated: sphere heat-kernel values \(a_6(S^2)=4/315\) , \(a_6(S^4)=74/63\) , \(a_6(S^6)=1139/63\) (round unit) and \(a_6(S^6)_{\rm conf}=5/63\) (conformal, correctly kept separate from the round-unit value), agreeing across two independent computation routes to \(\sim4\times10^{-14}\) ; the scale-free, Levi-Civita-immune ratio \(a_4/a_2^2=66/125\) ; the \(\mathbb{Z}_2\) -defect product-trace check on \(S^2\times(S^1_Y/\mathbb{Z}_2)\) , \((1/2)\cdot(4/315)=2/315\) (peeling to \(\sim1\times10^{-5}\) on one route, \(1.3\times10^{-14}\) on the Vandermonde route); and the twisted-circle trace \(\mathrm{Tr}(\sigma e^{-tD})=1.000000000000\ldots\) exactly, \(t\) -independent, with only integer powers appearing — which rules out the wrong-object framing "the boundary \(a_6\) needs a half-integer heat-kernel tower" that would otherwise stall the computation before it starts (published boundary-tower literature stops at \(a_5\) ; this object is a genuine extension, not a lookup).

 Why the object is honestly incomplete rather than merely unlooked-up: the canonical-connection Casimir/Peter–Weyl spectrum reproduces \(a_0\) and \(a_2\) exactly but is off by \(1/24\) at the vector level \(a_4\) — so the canonical-connection route cannot deliver a trustworthy \(a_6\) , and the Levi-Civita route is mandatory. On the Levi-Civita route, the graviton leg of \(a_6\) is blocked specifically at the Gelfand–Tsetlin off-diagonal hopping stratum: the hopping matrix elements mixing the five Weyl-inequivalent \(T^2\) weight classes are standard SU(3) GT lowering-operator formulas (square roots of products of pattern-entry differences), exact in principle, but not yet enumerated. This is a bounded computation-debt at a named stratum, not an in-principle gap — the certified shared core across both attempted routes (Gilkey/Lichnerowicz "Route A" and ghost+vector reconstruction "Route B") already includes the nine weight-6 curvature invariants, the full Lichnerowicz \(E_L\) spectrum, and \(a_0/a_2/a_4\) ; the scalar backbone ratio \(a_6/a_2^3=7936/39375\) is banked across three or more independent engines. Only the graviton hopping term itself remains OWED, and Route A and Route B have not yet reached two-route agreement on it.

 This object's status is OPEN/UNMADE — uncomputed, no sign asserted. Fabricating a total here (or assuming a convenient sign) would directly violate the fabrication guard; none is asserted. It is not the Clay wall: it is a finite computation on a compact fixed locus, not a continuum-uniform statement over infinite volume, and it is not dissolved by Granularity: a finite object at fixed cutoff is untouched by the P1 continuum reframe. Its correct route to closure is stated plainly: derive \(a_6^\partial\) target-blind \(\to\) finite \(c_3^\gamma\) \(\to\) evaluate \(P(a_6)\) ; a definite-sign violation of \(P(a_6)\ge0\) would itself be a legitimate refuting close, not a failure of method — the functional is a decision object (violation \(\Rightarrow\) refute; pass \(\Rightarrow\) consistency only, never a proof of the continuum wall). This object is shared with Gap-01 and is one of two UQF-3-local, non-Clay residuals carried forward unresolved, the other being R1 (the \(d=4\) boundary Chern–Simons/Dai–Freed/eta anomaly class, bulk-vanishing proven for \(d\le3\) by FOS Cor 7.5, but the \(d=4\) case OPEN and flagged likely nonzero by FOS Cor 7.6, exported to UQF-4).

 The frozen negative controls that any eventual \(a_6^\partial\) computation must reproduce or be provably wrong against, restated here at full precision: \(\dim K_6=6\) ; \(\mathrm{Ric}_i=5/12\) ; \(\mathrm{Scal}=5/2\) ; \(\mathrm{Scal}^2=25/4\) ; \(|\mathrm{Ric}|^2=25/24\) ; \(|\mathrm{Riem}|^2=23/12\) , ratio \(|\mathrm{Riem}|^2/\mathrm{Scal}^2=23/75=0.3066\overline6\) (never \(31/147\) — a retired Bianchi-inconsistent Nomizu-sign-bug artifact — and never \(60\) , the unrelated round- \(S^6\) value); \(|\mathrm{Ric}|^2/\mathrm{Scal}^2=1/6\) ; \(\mathrm{Scal}/\mathrm{Ric}_i=6=\dim K_6\) in both normalizations; cubic invariants \(K_1=-113/72\) , \(K_2=-5/72\) , \(|\nabla\mathrm{Riem}|^2=1/4\ne0\) (zero second-Bianchi violations — \(K_6\) is homogeneous but not locally symmetric, which is exactly why the \(a_6\) graviton leg carries the Gelfand–Tsetlin ladder term at all); \(\chi(K_6)=6\) , \(\chi(S^2)=2\) , \(\chi(S^1_Y/\mathbb{Z}_2)=1\) (exact topological); and the chirality index \(n_L=+3\) , \(n_R=0\) on \([0,\pi]\) . Separately, the graded fibre-weight data that must never be confused with this curvature ledger: on the 13-dimensional fibre with \(A=\mathrm{diag}(1_{12},-1)\) , \(\mathrm{tr}\,A=11\) , \(\mathrm{tr}\,A^2=13\) , graded graviton weight \(\mathrm{tr}\,\gamma_{\rm grav}=(\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2)/2=(13+121)/2=67\) , graded ghost weight \(\mathrm{tr}\,\gamma_{\rm ghost}=\mathrm{tr}\,A=11\) , Block-A graded weight \(=67-2(11)=45\) — the graded graviton weight \(67\) is never interchanged with the ungraded \(\mathrm{Sym}^2\) multiplet count \(91\) ; they are different quantities computed from the same \(A\) and answer different questions.

 I.5 The four Layer-2 admissibility screens

 Invariance. Cleanly satisfied at the level the gate actually certifies, and load-bearing. \(\mathcal H_{\rm phys}=\ker Q/\operatorname{im}Q\) is, by definition, the gauge-redundancy quotient — frame-independent with respect to the gauge group precisely because it is constructed to annihilate everything that is not. This is what makes "positive norm on \(\mathcal H_{\rm phys}\) " a coordinate-independent question at all, rather than an artifact of a gauge choice. Invariance here is inherited conditionally from UQF-4's \(Q^2=0\) nilpotency result, not re-derived inside UQF-3 — which is exactly why sub-claim (a), BRST no-ghost, is graded CERTIFICATE-CONDITIONAL rather than unconditionally DERIVED, and why the open \(d=4\) boundary anomaly class (R1) is exported to UQF-4 rather than assumed to cancel.

 Record-Interface. The finite-cutoff construction produces an actual, examinable physical record: the spectrum of \(H_a\) , the finite-volume gap \(\Delta(a,L)\) , the vacuum state singled out by Perron–Frobenius simplicity. These are computable, checkable numbers at any fixed \((a,L,\beta)\) — the theory is not merely asserted to be positive; it exposes a concrete self-adjoint operator whose spectrum can in principle be examined term by term. This is precisely the property underwriting the claim that no number is borrowed into the finite regime: every quantity in the banked leg is read directly off the record the finite construction produces, never imported from the unproven continuum.

 Causal-Order / target-blindness. The screen most directly at stake in a positivity gate, since positivity is exactly what licenses a sensible causal (Lorentzian) reconstruction downstream: reflection positivity is the Euclidean shadow of a self-adjoint Hamiltonian bounded below, the property that makes unitary time evolution \(e^{-itH}\) well-defined rather than exploding. Target-blindness is maintained throughout the construction: the proof never assumes the observed SM matter content is positive-norm and reverse-engineers a justification; it proves the Kugo–Ojima algebraic cancellation and the transfer-matrix bound structurally, with given-E entering only as the target list of what must come out positive, never biasing how the proof runs. The self-check is explicit in the residual ledger itself: R1, R4, R5 are carried as open inheritances/exports rather than assumed closed; R3 and R6 are named as walls that refuse to assume the answer; no residual anywhere in the roster is closed by reverse-engineering toward the desired positive-norm outcome.

 Nonseparability. Load-bearing against UQF-3 in the specific sense that it actively blocks the single most tempting shortcut: the Euclidean measure on the full 13-dimensional arena does not factorize as base(4D) \(\times\) internal( \(K_6\times S^2\times S^1_Y/\mathbb{Z}_2\) ). This is exactly why the naive move — "prove positivity on \(\mathcal M_4\) , prove it separately on the compact internal factors, then multiply" — is explicitly demoted from a prior "reduce" claim (retracted) to an unproven conjecture: residual R6, full KK tower + all-loop positive-definite inner product, DEMOTED to conjecture, blocked on R3 plus UQF-9/UQF-10, with no independent close available. Nonseparability is likewise the reason the GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM — that local positivity certificates on separate sectors automatically assemble into one global positive theory — is carried as a named, unproven theorem-debt , never a silently assumed lemma, and never presented as an axiom in disguise. Nonseparability does not merely bound how strong a claim can be made; it rules out the one simplification (factorize-and-multiply) that would otherwise make the continuum leg look closer to closed than the record supports.

 I.6 What the three roots jointly deliver

 Applied completely and jointly, the three deep roots produce the exact shape of UQF-3's terminal status. Shape supplies the target content and the geometric reflection map, and stops short of supplying uniqueness (W3 stays open, uninherited). Scale is a certified non-actor: zero of the four irreducible anchors are consumed, so there is no anchor-tuning route to a manufactured closure, honest or otherwise. Granularity does essentially all the positive work: it forces the finite-cutoff leg to DERIVED/ESTABLISHED status against published theorems (Osterwalder–Seiler, Lüscher), and it legitimately reduces the single continuum-uniform statement (SL-5/ULFCB \(\equiv\) R3) to an axiom-conditional idealization via P1 — a dissolution, not a solve, and one that is explicitly denied any license to also dissolve the genuinely finite R4 object, which survives open on its own terms. The four Layer-2 screens confirm the construction is honest rather than smuggled: Invariance is correctly conditioned on an external, named, audited input; Record-Interface shows the finite leg is a real, examinable operator spectrum and not a bare assertion; Causal-Order/target-blindness is maintained because given-E never leaks from "target list" into "certificate"; and Nonseparability is load-bearing against the gate, actively forbidding the factorize-and-multiply shortcut and forcing the composition theorem to be carried as named theorem-debt. Together these fix the terminal precisely: a real, DERIVED finite-cutoff result, a target-blind SCOPE theorem separating this gate from the mass-gap problem, one continuum-uniform wall correctly dispositioned REDUCED-TO-AXIOM under granularity, and one genuinely finite open computation (R4/ \(a_6^\partial\) ) that the roots correctly refuse to either fabricate or prematurely dissolve.

 Construction II - the full derivation

 II.0 What this section derives, and how it divides the labor

 UQF-3 has two sub-conditions on one object — (a) BRST no-ghost, \(\mathcal H_{\rm phys}=\ker Q/\operatorname{im}Q\) carries only non-negative-norm states, and (b) Euclidean reflection positivity, which via Osterwalder–Schrader reconstruction supplies a self-adjoint, bounded-below Hamiltonian with a genuine ground state. The finite-cutoff proof of (b) — the transfer-matrix identity \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) , Perron–Frobenius vacuum, strict gap \(\Delta(a,L)>0\) — is carried out in full in the next construction and is not repeated here. This section derives everything else that stands between "the finite-cutoff Euclidean theory is positive" and "the theory is positive": the retained-sector BRST/Kugo–Ojima quartet mechanism (sub-claim (a)), the free-field bridge that shows (a) and (b) are the same statement in the regime where both are under control, the graviton's own positivity check on the frozen internal geometry, the SCOPE theorem that proves UQF-3 is a strictly smaller problem than the Gap-02 mass gap, and the continuum spine SL-0 through SL-5 that localizes every piece of open content down to one named, irreducible analytic statement. Every object below is pinned at all three layers of the frozen branch; no step introduces new geometry, and no constant is asserted without either a proof or an explicit OPEN flag.

 II.1 Sub-claim (a): the BRST/Kugo–Ojima quartet mechanism, in full

 The three-layer object. The gauge-fixed continuum sector sits on: \(\times\) Stage — the retained 4D zero-mode sector of \(\mathcal M_4\) tensored against the low-lying Kaluza–Klein tower of \(K_6\times S^2\times S^1_Y/\mathbb Z_2\) , with gauge group \(SU(3)_c\times SU(2)_L\times U(1)_Y\) read off the isometries of \(K_6\) , \(S^2\) , \(S^1_Y\) respectively (mod the \(\mathbb Z_6\) center identification, Smith normal form invariant factors \([1,6,6]\) ). \(\oplus\) Rulebook — covariant (BRST) gauge-fixing is imposed here, in contrast to the Wilson-kinematics leg of sub-claim (b): a gauge-fixing term \(\tfrac1{2\xi}(\partial^\mu A_\mu^a)^2\) plus Faddeev–Popov ghosts \((\bar c^a,c^a)\) are added to the Lagrangian, together with the BRST grading (ghost number) and the nilpotent BRST differential \(Q\) , \(Q^2=0\) (audited input, exported from and to UQF-4; not re-derived here — counting it as a UQF-3 floor anchor would be a relabel error). \(\otimes\) Actors — the BRST charge \(Q_{\rm BRST}\) acting on the off-shell Fock space, the ghost/antighost creation-annihilation algebra, and the readout \(\mathcal H_{\rm phys}=\ker Q/\operatorname{im}Q\) .

 The quartet mechanism, mode by mode. For each momentum mode of the gauge field \(A_\mu^a\) , covariant gauge-fixing produces four propagating "would-be" degrees of freedom instead of the two physical transverse polarizations: the two transverse modes \(A_\mu^{a,\perp}\) , the longitudinal mode \(A_\mu^{a,\parallel}\) , and the timelike/scalar mode \(A_0^a\) . Alongside these, the ghost sector supplies one complex pair \((c^a,\bar c^a)\) per mode. The BRST transformation acts as
$$
Q\,A_\mu^a = \partial_\mu c^a + \dots,\qquad Q\,c^a = -\tfrac12 f^{abc}c^bc^c,\qquad Q\,\bar c^a = B^a\ (\text{Nakanishi–Lautrup field}),\qquad Q\,B^a=0,
$$
with \(B^a\) the auxiliary field enforcing the gauge condition on-shell ( \(B^a=\partial^\mu A_\mu^a/\xi\) after using its own equation of motion in Feynman-type gauges). The Kugo–Ojima quartet theorem (Kugo–Ojima, 1979) organizes the longitudinal mode \(A_\mu^{a,\parallel}\) , the timelike mode \(A_0^a\) , the ghost \(c^a\) , and the antighost \(\bar c^a\) into a single BRST quartet: \(Q\) maps
$$
\bar c^a \ \xrightarrow{Q}\ B^a\ \xrightarrow{Q}\ 0, \qquad (\text{longitudinal/timelike gluon combination}) \ \xrightarrow{Q}\ (\text{ghost combination}) \ \xrightarrow{Q}\ 0,
$$
i.e. each quartet member is either \(Q\) -exact (lies in \(\operatorname{im}Q\) ) or maps to a \(Q\) -exact partner. A standard cohomological lemma (the "quartet mechanism" lemma of Kugo–Ojima, restated in Weinberg QFT Vol. II §15.8 and Schwartz QFT and the Standard Model Ch. 25 for the textbook, free-field case) then shows that the entire quartet decouples from \(\mathcal H_{\rm phys}=\ker Q/\operatorname{im}Q\) : any state built by acting with the longitudinal/timelike gluon or ghost/antighost creation operators on the vacuum is either not \(Q\) -closed (so does not enter \(\ker Q\) ) or is \(Q\) -exact (so is quotiented out by \(\operatorname{im}Q\) ). What survives the cohomology is exactly the two transverse gluon polarizations per momentum mode, matching the physical, gauge-invariant counting.

 Sign structure and why it is not automatic. The quartet argument is not merely a counting exercise — the reason it delivers positive -norm survivors rather than a wash is the specific indefinite metric structure of the timelike mode paired against the ghost sector. In covariant quantization the timelike photon/gluon mode \(A_0^a\) carries a negative -norm one-particle state under the naive Fock inner product; this negative norm is exactly cancelled, mode by mode, by the ghost pair, whose anticommuting statistics flips the sign of their contribution to the physical-subspace projector. The quartet mechanism is precisely the statement that this cancellation is exact and universal (holds order by order, not just at tree level, given \(Q^2=0\) ), so that the residual — what actually survives into \(\ker Q/\operatorname{im}Q\) — carries strictly non-negative norm. This is why sub-claim (a) is a genuine positivity statement and not a bookkeeping identity: a theory in which the ghost sector had the wrong statistics, or in which \(Q^2\ne0\) so that \(\operatorname{im}Q\not\subset\ker Q\) , would not have this cancellation, and \(\mathcal H_{\rm phys}\) as naively defined would fail to be well-posed at all (the quotient would not make sense, since \(Q\) would not be a differential). The entire construction rests on \(Q^2=0\) , which is why sub-claim (a) is graded CERTIFICATE-CONDITIONAL rather than unconditionally DERIVED: it is conditional on the audited but externally-verified fact that the frozen branch's BRST operator is nilpotent, a result whose own open content (the single even-degree \(d=4\) boundary Chern–Simons/Dai–Freed/eta-class anomaly on \(S^1/\mathbb Z_2\) and on \(K_6\times S^2\) ) is tracked as residual R1 and exported to UQF-4, not re-litigated here. The bulk vanishing of this anomaly class is proven for \(d\le3\) (FOS Corollary 7.5); the \(d=4\) case relevant here is open and flagged likely nonzero (FOS Corollary 7.6) — an honest, named gap, not assumed away.

 Matter and Higgs sectors. Away from the gauge quartet, positivity is immediate by construction rather than by cohomological cancellation. The matter bundle \(\mathcal E_{\rm matter}=S_{3,1}\otimes S^{\rm spin^c}_{K_6}\otimes S^{\rm spin^c}_{S^2}\otimes L_Y\otimes V_{SU(3)}\otimes V_{SU(2)}\otimes V_{F^+}\) carries the standard positive-definite Dirac inner product on each spinor factor (the spin- \(\mathbb C\) structure on \(K_6\) fixes the family index to \(\chi(K_6,E)=-3\) , three chiral generations, with chirality projector \(P_\chi=\tfrac12(1+\gamma_5\Gamma_8)\) and Atiyah–Singer–Patodi index \(n_L=+3\) , \(n_R=0\) on the active interval \([0,\pi]\) — no surviving mirror partner to threaten a wrong-sign norm), and none of the matter fields are subject to gauge-fixing indefiniteness since they carry no gauge-redundant components. The Higgs sector \(\mathcal E_{\rm Higgs}=L_\gamma\otimes V_{SU(2),\rm doub}\) , realized as a Wilson-line/Hosotani mode with integer winding \(n_H=1\) , is likewise a genuine positive-definite doublet with no compensating ghost required, since the Wilson-line holonomy is not itself a gauge-fixed field requiring an FP determinant. Both sectors are therefore positive-norm unconditionally , not merely conditionally on \(Q^2=0\) ; only the gauge sector's positivity depends on the quartet cancellation above.

 II.2 The free-field OS \(\leftrightarrow\) BRST equivalence — where sub-claims (a) and (b) meet

 Sub-claims (a) and (b) are, at the free (non-interacting) level, provably the same statement viewed through two different quantization schemes — this is a standard textbook equivalence (Osterwalder–Schrader reconstruction of a free/perturbative gauge theory reproduces the BRST-quotient Hilbert space exactly), and it is the bridge that lets the finite-cutoff Wilson-kinematics leg (sub-claim (b), no gauge-fixing, no ghosts) and the covariant BRST leg (sub-claim (a), gauge-fixed, ghosts present) be recognized as two routes to the same physical content rather than two independent, potentially inconsistent, claims about the theory.

 The argument runs as follows. In the free-field limit (coupling \(g\to0\) , or equivalently the quadratic part of the gauge action alone), the Wilson-kinematics path integral of Construction III's Step 1–4 can be gauge-fixed after the fact by inserting a resolution of the identity over gauge orbits, without altering any gauge-invariant correlator — this is legitimate precisely because at \(g=0\) the Faddeev–Popov determinant is a field-independent (orbit-volume) constant, so introducing it changes nothing about which states have positive norm. Doing so converts the manifestly reflection-positive, ghost-free Euclidean measure into the covariant, BRST-gauge-fixed Euclidean measure with ghosts, and the Osterwalder–Schrader reconstruction theorem applied to the latter produces a Hilbert space that is unitarily equivalent to \(\ker Q/\operatorname{im}Q\) built directly from the BRST-quantized free theory (this equivalence is worked through explicitly in Weinberg QFT Vol. II Ch. 15 and, for the reflection-positivity side specifically, is the free-field case the interacting Osterwalder–Seiler construction generalizes). Concretely: the free transverse gluon two-point function computed from the gauge-invariant Wilson correlator agrees exactly with the free transverse gluon two-point function computed from the BRST-quotient Fock space, because both are fixed uniquely (up to the overall positive normalization) by Poincaré invariance, masslessness, and the transverse-projector numerator \(\eta_{\mu\nu}-k_\mu k_\nu/k^2\) — there is no room for the two constructions to disagree at the free level since the physical (gauge-invariant) content is unique.

 Scope of the equivalence and why it must not be over-extended. This equivalence is explicitly free-field-scoped — it is a fact about the quadratic action, where the Faddeev–Popov determinant is field-independent and the gauge-fixing procedure commutes cleanly with the reflection-positive completion. At any finite coupling \(g\ne0\) , the Faddeev–Popov determinant becomes field-dependent (a functional of the gauge connection), and inserting the gauge-fixing resolution of the identity into the interacting Osterwalder–Seiler measure is no longer a trivial reweighting — it is exactly the step that would need to be controlled to extend sub-claim (a)'s conditional, perturbative statement into a nonperturbative, interacting-theory statement, and this control is not established. Presenting the free-field textbook equivalence as if it validated the interacting branch would be precisely the fabrication pattern the brief flags and forbids ("free-field textbook citation proves the interacting branch" is a named non-claim). What the free-field equivalence does legitimately establish is that sub-claims (a) and (b) are not two independent risks stacked on top of each other in the regime where both are controlled — they are the same underlying positivity fact, checked twice by two different, mutually consistent methods, which is a nontrivial and useful cross-validation, not a proof that either method's domain of validity extends further than it does.

 II.3 Linearized graviton positivity — the transverse-traceless check on the frozen internal geometry

 The metric perturbation \(h_{\mu\nu}\) around the 13-dimensional background decomposes, after the standard TT (transverse-traceless) gauge-fixing, into the physical graviton polarizations plus pure-gauge and constrained pieces. The relevant frozen-geometry object is the bundle \(\mathrm{Sym}^2_0\,T^*K_6\) (the traceless symmetric 2-tensor bundle on the internal factor, dimension 20 as a real vector bundle at each point, since the full symmetric-square bundle \(\mathrm{Sym}^2\,T^*K_6\) has dimension \(\binom{6+1}{2}=21\) and removing the trace leaves \(21-1=20\) ), equipped with the Lichnerowicz Laplacian endomorphism
$$
(E_L h) {ab} = \mathrm{Ric} {ac}h^c{} b + \mathrm{Ric} {bc}h^c{} a - 2R {acbd}h^{cd}.
$$
At the Einstein center of the Killing-form normal metric ( \(\mathrm{Ric}=\tfrac5{12}g\) ), this operator's spectrum on the full \(\mathrm{Sym}^2\,T^*K_6\) (dimension 21) is exactly
$$
\Big{\ \tfrac16\,(\times6),\ \ \tfrac5{12}\,(\times6),\ \ \tfrac76\,(\times6),\ \ \tfrac{17}{12}\,(\times2),\ \ \tfrac53\,(\times1,\ \text{pure-trace mode})\ \Big},
$$
and restricting to the transverse-traceless subbundle \(\mathrm{Sym}^2_0\) (dimension 20) removes the pure-trace eigenvalue \(5/3\) (mode count \(21\to20\) ), leaving
$$
\mathrm{spec}(E_L)\big|_{\mathrm{Sym}^2_0} = \Big{\ \tfrac16\,(\times6),\ \ \tfrac5{12}\,(\times6),\ \ \tfrac76\,(\times6),\ \ \tfrac{17}{12}\,(\times2)\ \Big},
$$
with traces \(\mathrm{tr}\,E_L=\tfrac{40}3\) and \(\mathrm{tr}\,E_L^2=\tfrac{241}{18}\) over this 20-dimensional subbundle (both exact rationals, cross-checked directly against the eigenvalue list: \(6(\tfrac16)+6(\tfrac5{12})+6(\tfrac76)+2(\tfrac{17}{12}) = 1+\tfrac52+7+\tfrac{17}6 = \tfrac{6+15+42+17}6=\tfrac{80}6=\tfrac{40}3\) , confirming \(\mathrm{tr}\,E_L=40/3\) ; similarly \(6(\tfrac1{36})+6(\tfrac{25}{144})+6(\tfrac{49}{36})+2(\tfrac{289}{144}) = \tfrac16+\tfrac{25}{24}+\tfrac{49}6+\tfrac{289}{72}\) , which over the common denominator 72 is \(\tfrac{12+75+588+289}{72}=\tfrac{964}{72}=\tfrac{241}{18}\) , confirming \(\mathrm{tr}\,E_L^2=241/18\) ). Every eigenvalue in this list is strictly positive: \(1/6, 5/12, 7/6, 17/12\) are all manifestly \(>0\) .

 What this establishes and what it does not. A strictly positive Lichnerowicz spectrum on \(\mathrm{Sym}^2_0\,T^*K_6\) means the graviton kinetic operator on the internal factor has no zero modes and no wrong-sign (tachyonic or ghost-like) directions at the linearized level restricted to the internal geometry's contribution — this is exactly the statement needed to confirm that the 4D graviton, after Kaluza–Klein reduction, does not inherit a negative-norm polarization from the compactification. Combined with the standard flat-space TT counting on \(\mathcal M_4\) (a massless spin-2 field in four dimensions carries exactly 2 physical transverse-traceless polarizations, the textbook linearized-gravity result reproduced by any consistent decomposition of \(h_{\mu\nu}\) into scalar, vector, and TT-tensor pieces via the Stewart–Walker/York decomposition, with the scalar and vector pieces removed by diffeomorphism gauge-fixing exactly as the longitudinal/timelike gluon modes are removed by the Yang–Mills BRST quartet), this fixes the positive-norm graviton content at 2 polarizations, matching observation. This result is graded DERIVED-GIVEN-E (linearized) : it is a genuine, checked positivity fact about the frozen internal geometry's contribution to the graviton kinetic term, but it is explicitly restricted to the linearized (quadratic-in- \(h\) ) theory. Nonlinear graviton self-interactions and the full KK tower of graviton modes above the zero mode are not certified positive by this computation — that is residual R7 , carried forward as OPEN above the cutoff and exported to UQF-14, since certifying it requires a UV completion of gravity that this gate does not supply (a UQF-9-blocked dependency, not a UQF-3 failure).

 II.4 The SCOPE theorem: positivity is a proper sub-wall of the mass gap, never the reverse

 The single genuinely new, target-blind logical move UQF-3 contributes — as opposed to importing theorems — is a scope result relating this gate to the Gap-02 mass-gap problem. The claim is precise: the finite-cutoff and free-field positivity results above require none of the mass-gap machinery , while the reverse dependency (mass-gap arguments requiring positivity as an input) is standard and unavoidable. Concretely:

 Construction III's transfer-matrix identity \(T=e^{-aH_a}\) , \(H_a\ge0\) , and the Perron–Frobenius vacuum-uniqueness argument use only: (i) gauge-invariance of the Wilson action, (ii) reflection symmetry of the lattice measure, (iii) strict positivity of the Boltzmann weight on the interior of the compact configuration space. None of these three ingredients references a mass gap, a confinement scale, or any dynamically generated scale \(\Lambda_{\rm YM}\) — the argument goes through identically whether or not the theory happens to be gapped, confining, or even asymptotically free.

 By contrast, every standard proof strategy for a mass gap (the SL-1 clustering argument, the SL-5 large-field-domination bound) is stated conditional on a positive-definite Hilbert space with a self-adjoint, bounded-below Hamiltonian already in hand — one cannot even pose the question "is there a spectral gap above the vacuum" without first having a Hilbert space on which \(\mathrm{Spec}(H)\) is a well-defined subset of \(\mathbb R_{\ge0}\) . Positivity is a logical prerequisite for the mass-gap question to be well-posed, not a consequence of it.

 This proves a strict one-way dependency,
$$
(\text{gap-machinery}) \ \longrightarrow\ (\text{positivity}), \qquad (\text{positivity})\ \not\longrightarrow\ (\text{gap-machinery}),
$$
and therefore that UQF-3's positivity requirement is a proper sub-wall of the Gap-02 mass-gap problem , not a restatement of it and not equivalent to it. This is why UQF-3 can reach a terminal grade of its own (CERTIFIED-IRREDUCIBLE / RESOLVED +0) at the finite-cutoff level even while Gap-02's continuum mass gap remains an open Clay-class problem: UQF-3's finite-cutoff leg is unconditionally established without needing Gap-02 solved, and the only place the two problems become the same object is at the single continuum-uniform wall identified below (SL-5/R3), where both gates are independently forced to confront the identical ULFCB obstruction. The scope theorem is what licenses treating that shared wall as "this is provably the same difficulty the whole field faces" rather than "this program has an unusually hard private problem" — the direction of implication is proven, not asserted.

 II.5 The continuum spine, SL-0 through SL-5 — localizing all remaining open content to one statement

 The full logical chain from "finite-cutoff positivity is established" to "the continuum theory is positive" is organized as six sub-lemmas. Each is stated precisely, graded honestly, and — critically — the chain is constructed so that all uncontrolled content is funneled into the single last link, rather than being smeared indeterminately across several vague steps.

 SL-0 (variational identity — banked, no open content). For any self-adjoint \(H\) bounded below with candidate vacuum \(\Omega\) , the following are logically equivalent by the spectral theorem alone: \(\inf\mathrm{Spec}(H)|_{\mathcal H\ominus\Omega}>0\) ; there is no sequence of unit vectors \(\psi_n\perp\Omega\) with \(\langle\psi_n,H\psi_n\rangle\to0\) (no "soft sequence"); \(H\) is coercive off \(\Omega\) . This is a restatement of the min-max characterization of the spectrum and holds for any Hilbert space and any such \(H\) — pure functional analysis, theory-independent, banked in full.

 SL-1 ( \(\mathcal M_4\) spectral conversion — banked, conditional structure only, no fabricated constant). If the Euclidean two-point function of a local reflection-positive operator \(\mathcal O\) clusters uniformly, \(|\langle\mathcal O(x)\mathcal O(0)\rangle-\langle\mathcal O\rangle^2|\le Ce^{-\Delta|x|}\) as \(|x|\to\infty\) , then , via the standard OS-reconstruction spectral representation, \(\mathrm{Spec}(H)\cap(0,\Delta)=\emptyset\) with \(\Delta\ge c'\Lambda_{\rm YM}\) for some structure constant \(c'>0\) and a dynamically generated scale \(\Lambda_{\rm YM}\) . This step derives the logical structure linking real-space clustering to a spectral gap; it does not supply a numeric value for \(c'\) , and none is fabricated here — \(c'\) is a placeholder for "some positive constant fixed by the clustering rate," not a computed number.

 SL-2 (finite- \(a\) base — banked/established). This is exactly Construction III in full: \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) self-adjoint, Perron–Frobenius-simple vacuum, strict finite-volume gap \(\Delta(a,L)>0\) , with an independent analytic lower bound in the strong-coupling regime \(\beta<\beta_{\rm conv}\) . This supplies the concrete Hilbert space and Hamiltonian that SL-0's variational identity is applied to .

 SL-3 (continuum kernel construction — OPEN, the prior question). Does \(\mathcal H_{\rm phys}^{(a,L)}\) , equipped with its positive-definite reflection-positive inner product, converge as \(a\to0,L\to\infty\) to a genuine, still-positive-definite Hilbert space \(\mathcal H_{\rm phys}^{\rm YM}\) ? This is logically prior to asking whether that limiting space has a mass gap: one must first know a sensible limiting Hilbert space exists at all. The named obstruction is that every known continuum construction of Yang–Mills that proceeds via gauge-fixing runs into the Gribov ambiguity as \(a\to0\) (Gribov 1978; Singer 1978, the topological no-go showing no continuous global gauge-fixing exists on the space of connections modulo gauge; Neuberger 1987, showing the naive BRST/FP continuum path integral vanishes identically, \(0/0\) , once the Gribov copies are correctly summed with signs) — precisely the ambiguity the finite-cutoff Wilson-kinematics leg of sub-claim (b) sidesteps by using no gauge-fixing at all. Whether a gauge-fixing-free (Wilson-kinematics) continuum limit avoids this obstruction, and if so whether it still converges to a positive-definite kernel, is the open content of SL-3. No claim is banked here beyond the finite- \(a\) statement of SL-2.

 SL-4 (Wightman/OS axioms H1–H3 — OPEN, Clay-class). Assuming SL-3's kernel exists, does it satisfy the remaining Osterwalder–Schrader/Wightman axioms needed for a physical quantum field theory: existence of the continuum measure, survival of reflection positivity through the limit, and nontriviality (the theory is not a free/Gaussian theory in disguise, and correlators do not degenerate)? This is exactly the analytic content of the Yang–Mills existence half of the Clay Millennium problem statement, and it is open in the identical sense the Clay problem is open — no proof, no disproof, for any interacting 4D non-abelian gauge theory.

 SL-5 (H4 / ULFCB — OPEN, the wall itself; no constants fabricated). Even granting SL-3 and SL-4, the mass-gap half of the Clay statement (and, more immediately for UQF-3, the uniform positivity/coercivity statement that would let SL-1's clustering argument be run uniformly through the continuum limit) requires a uniform large-field coercivity bound (ULFCB):
$$
\inf{\langle\psi,H\psi\rangle} \;\ge\; c'\,\Lambda_{\rm YM} \;>\;0 \qquad \text{uniformly in the continuum/OS limit},
$$
equivalently a uniform link-measure bound \(\|\Phi_n\|_\kappa\le K\) for all RG scales \(n\) , all \(a\) , all \(L\) , all \(\beta\) . No constants \(c',K,\kappa\) are supplied — supplying them without a proof would be fabrication, and none is given. The precise obstruction, as far as it is understood, is due to Bałaban's rigorous renormalization-group analysis (CMP, roughly 1984–1989): at RG step \(n\) , the measure of the "large-field" (badly-behaved) region satisfies a Gaussian-type suppression bound \(\mu_n(L_n)\le e^{-c/g(2^na)^2}\) , showing large-field configurations are rare . But rarity is not domination : ULFCB requires that the rare large-field directions be structurally incapable of hosting a soft, vacuum-orthogonal sequence in the SL-0 sense — a stronger, structural claim that rarity alone does not deliver. Marginal (asymptotically free but not superrenormalizable) \(d=4\) Yang–Mills is exactly the borderline case with no spare coercive margin : the coupling runs logarithmically rather than being suppressed by a positive power of the cutoff, so there is no perturbative small parameter left over to convert "rare" into "dominated" by a simple estimate. This is the single wall the whole chain funnels into. It is Bałaban's own scoped result (UV stability of the RG flow) that is imported here — his construction controls the UV continuum limit's existence step by step, but explicitly does not by itself supply the continuum Hilbert space construction, the reflection positivity survival, or the mass gap (a standing misattribution risk flagged explicitly: citing Bałaban's UV-stability results as if they closed SL-3–SL-5 would overclaim what his theorems prove).

 How the six steps compose. SL-0 is unconditional. SL-1 is a conditional bridge (structure only). SL-2 is the fully banked finite-cutoff achievement (Construction III). SL-3, SL-4, and SL-5 are the three progressively deeper open questions that stand between the finite-cutoff result and the continuum Clay-class statement, with SL-5/ULFCB identified as the deepest — the one all of Gap-02's and UQF-3's continuum content reduces to. This is precisely the content of residual R3 : "constructive positivity for the full interacting 4D base," dispositioned REDUCED-TO-AXIOM / precisely-OPEN / CERTIFIED-IRREDUCIBLE — identical in kind to the 4D Yang–Mills Clay existence/positivity problem, with no lever inside this framework (or, so far as is publicly known, inside any framework) to close it. The granularity axiom P1 (Construction I, §I.3) is what licenses treating this specific, named, irreducible gap as a limit on all present knowledge rather than a defect unique to this construction: P1 asserts the finite-cutoff theory (SL-0 through SL-2, fully banked) is the physically relevant object, and the SL-3–SL-5 continuum question is reduced to an axiom-conditional idealization rather than counted as an unmet physical requirement.

 II.6 What Construction II establishes, stated without hedging

 Collecting the five results of this section: (1) the BRST/Kugo–Ojima quartet mechanism proves sub-claim (a) — the retained gauge sector's would-be ghosts (longitudinal/timelike gluon plus ghost/antighost) cancel exactly, leaving 2 positive-norm transverse polarizations per mode, conditional only on the externally-audited \(Q^2=0\) and with matter/Higgs unconditionally positive-norm; (2) the free-field OS↔BRST equivalence proves sub-claims (a) and (b) are the same fact, checked by two independent methods, in the regime (free/quadratic) where both are under control, with no overreach into the interacting regime; (3) the linearized graviton computation on the frozen \(\mathrm{Sym}^2_0\,T^*K_6\) bundle proves the internal geometry contributes no wrong-sign polarization to the 4D graviton at the quadratic level, fixing exactly 2 positive-norm graviton polarizations, with nonlinear/KK-tower content honestly carried forward as open (R7); (4) the SCOPE theorem proves, by exhibiting the actual logical dependencies used in each direction, that positivity is a proper sub-wall of the mass gap and not the reverse, licensing UQF-3's independent terminal grade; and (5) the SL-0–SL-5 spine proves that every remaining piece of open content — SL-3's continuum kernel question, SL-4's Wightman/OS axioms, and SL-5's uniform coercivity bound — funnels into one named, irreducible statement (ULFCB, coextensive with the Clay mass-gap wall), rather than being scattered across an unbounded list of vague uncertainties. Nothing here fabricates a constant, assumes a target value, or treats a finite computation as dissolved by the continuum axiom P1: the finite \(a_6^\partial\) boundary object of Construction I §I.4 remains separately tracked as OPEN precisely because it is not touched by any step in this section.

 Construction III - the central result at full precision

 What this section proves, precisely. UQF-3 turns on a single boxed operator identity plus its four positivity corollaries, established unconditionally on the complete frozen 13-dimensional branch at fixed lattice spacing \(a>0\) and finite volume \(L<\infty\) :
$$
\boxed{\ T \;=\; e^{-aH_a}\,,\qquad 0\ \le\ T\ \le\ 1\,,\qquad H_a \;=\; H_a^{\dagger} \;\ge\; 0\ }
$$
together with Perron–Frobenius simplicity of the vacuum and a strictly positive finite-volume spectral gap \(\Delta(a,L)>0\) — and, sitting immediately above this banked identity, its exact logical ceiling: the identical positivity statement is not established uniformly through \(a\to0,\,L\to\infty\) for the fully interacting four-dimensional base. Below, every step of the finite-cutoff derivation is shown in full with no step skipped, the six-rung sub-lemma ladder (SL-0…SL-5) that localizes the open content to one named analytic statement is walked rung by rung, the one genuinely new logical result of the gate (the scope theorem separating UQF-3 from Gap-02) is derived, and the exact-rational geometric cross-checks feeding the not-yet-closed boundary functional \(P(a_6)\) are assembled and independently checked against frozen negative controls. No number below is fitted to a desired outcome, and the finite-cutoff sector — the part of this construction that is unconditionally derived — consumes zero dimensionful anchors.

 III.1 The three-layer object the theorem is proved on

 The transfer matrix is built on the complete frozen branch, never on a metric-only slice of it; every layer is load-bearing for the positivity argument that follows.

 \(\times\) Stage. The full arena is
$$
\mathfrak B_{\rm active} \;=\; \mathcal M_4 \times K_6 \times S^2 \times S^1_Y/\mathbb Z_2, \qquad K_6 = SU(3)/T^2 \ (\text{the } A_2 \text{ flag manifold}), \qquad D = 4+6+2+1 = 13.
$$
The Euclidean time direction singled out for reflection is one coordinate of \(\mathcal M_4\) ; the reflection \(\tau\mapsto-\tau\) acts on that slice. The internal directions \(K_6\times S^2\times S^1_Y/\mathbb Z_2\) ride along as spectator fibres at every lattice site, contributing the compact gauge group and matter content but carrying no transfer-matrix time direction of their own. \(S^1_Y/\mathbb Z_2\) carries its own, separate \(\mathbb Z_2\) reflection \(\theta\mapsto-\theta\) with two isolated fixed points at \(\theta=0,\pi\) ; the Donnelly equivariant defect there,
$$
\sum_{\rm fixed\ pts} \frac{1}{|1-dg|} \;=\; 2\times\frac{1}{|1-(-1)|} \;=\; 2\times\frac12 \;=\; 1,
$$
is a structurally distinct reflection from the Euclidean time-reflection that drives \(T=e^{-aH_a}\) below — it is the reflection that instead feeds the still-open boundary heat-kernel object \(a_6^\partial\) of §III.6, not the finite-cutoff Hamiltonian construction of §III.2. Keeping these two reflections (one on \(\mathcal M_4\) , one on \(S^1_Y/\mathbb Z_2\) ) distinct is itself a load-bearing bookkeeping fact for this gate: conflating them would misattribute the banked SL-2 result to the open R4 object or vice versa.

 \(\oplus\) Rulebook. Gauge-invariant Wilson lattice kinematics: link variables \(U_\ell \in SU(3)_c\times SU(2)_L\times U(1)_Y\) (identified modulo the global \(\mathbb Z_6\) center, Smith normal form invariant factors \([1,6,6]\) , \(G_{\rm SM}=(SU(3)_c\times SU(2)_L\times U(1)_Y)/\mathbb Z_6\) ) on a hypercubic lattice of spacing \(a\) , plaquette gauge action, Wilson/staggered matter action — and, critically, no gauge-fixing at any stage of this construction . This single rulebook choice is what structurally forecloses the Gribov ambiguity (Gribov 1978; Singer 1978; sharpened by Neuberger 1987): there is no gauge slice being cut, hence no non-uniqueness of the slice to obstruct a positive kernel. The granularity axiom P1 is the second load-bearing rulebook element: it is what licenses treating finite \((a,L)\) as the physically relevant regime that this section's theorem is about , rather than as a mere calculational stepping-stone to a continuum limit that is separately and honestly carried as open (§III.4, SL-3–SL-5).

 \(\otimes\) Actors. The link-reflection operator \(\theta: U_\ell \mapsto U_{\theta\ell}^{\dagger}\) (Euclidean time-reflection combined with Hermitian conjugation of the reflected links), the transfer operator \(T\) built from the plaquette and matter Boltzmann weights, and the physical Hilbert space \(\mathcal H_{\rm phys}^{(a,L)}\) obtained as the completion of the pre-Hilbert space of reflected-doubled observables under the reflection-positive inner product defined in Step 1 below. The endomorphism content entering the matter sector is drawn from the standard tensor ledger \(\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\) ; the readout of the whole construction is the spectrum of \(H_a = -\tfrac1a\log T\) .

 III.2 The finite-cutoff construction, every step shown

 Step 1 — the pre-Hilbert space and the reflection. On the Euclidean lattice with time-slicing at spacing \(a\) , let \(\Omega^+\) denote the algebra of observables built from link and matter variables supported at times \(\tau\ge0\) , and let \(\theta\) be the time-reflection \(\tau\mapsto-\tau\) combined with Hermitian conjugation on the link/matter fields — the Osterwalder–Seiler prescription for lattice gauge theory (Ann. Phys. 110 (1978) 440), extending the original continuum Osterwalder–Schrader construction (CMP 31 (1973) 83; CMP 42 (1975) 281) to compact-group lattice variables. Define the sesquilinear form
$$
\langle f,g\rangle \;\equiv\; \big\langle\, \theta(f)\, g \,\big\rangle_{\rm lattice}, \qquad f,g\in\Omega^+,
$$
where \(\langle\cdot\rangle_{\rm lattice}\) is the finite-volume, finite- \(a\) Euclidean path-integral expectation with the plaquette gauge action and Wilson/staggered matter action — a manifestly convergent finite-dimensional integral, because the gauge group \(SU(3)\times SU(2)\times U(1)\) is compact at each of the finitely many links on a finite lattice.

 Step 2 — reflection positivity (Osterwalder–Seiler theorem). Because the action is built entirely from gauge-invariant plaquette and Wilson-loop variables — no gauge-fixing term, no Faddeev–Popov ghost determinant, no BRST data anywhere in this construction — the Osterwalder–Seiler theorem applies without modification:
$$
\langle f,f\rangle \;=\; \big\langle\,\theta(f)\,f\,\big\rangle_{\rm lattice} \;\ge\; 0 \qquad \forall\, f\in\Omega^+.
$$
The mechanism is elementary and worth making fully explicit rather than citing as a black box: the plaquette weight \(\exp\!\big(-a^4\sum_\square \tfrac{1}{2g^2}\,{\rm Re\,Tr}(1-U_\square)\big)\) factorizes, across the \(\tau=0\) time-slice, into a product of a "positive-time" factor built from links at \(\tau\ge0\) and its \(\theta\) -image built from links at \(\tau\le0\) , and the Haar measure on each compact gauge-group factor at each link is reflection-symmetric. Hence \(\theta\) is an anti-unitary involution compatible with the measure — exactly the Osterwalder–Schrader positivity input, transplanted verbatim to the lattice. No continuum limit, no gauge-fixing choice, and no restriction on the coupling \(g\) is used anywhere in this step: the inequality holds at every finite \(a\) , every finite \(L\) , and every coupling.

 Step 3 — quotient and completion: the physical Hilbert space. The form \(\langle\cdot,\cdot\rangle\) is positive semi-definite but may carry a null space \(\mathcal N=\{f: \langle f,f\rangle=0\}\) . The physical Hilbert space is the quotient completion
$$
\mathcal H_{\rm phys}^{(a,L)} \;\equiv\; \overline{\Omega^+/\mathcal N}^{\ \langle\cdot,\cdot\rangle}.
$$
By construction every nonzero vector in \(\mathcal H_{\rm phys}^{(a,L)}\) has strictly positive norm: this is the operational content of "no ghosts" at finite cutoff. The quotient has already discarded every null direction in Step 3, and Step 2's inequality guarantees there is no negative -norm direction left over for the quotient to have missed — the two possibilities (null, and strictly positive) exhaust the sign spectrum of a positive-semi-definite form, so nothing negative can survive.

 Step 4 — the transfer operator, and the boxed identity. Time-translation by one lattice unit \(a\) induces a linear operator \(T\) on \(\mathcal H_{\rm phys}^{(a,L)}\) via
$$
\langle f, Tg\rangle \;=\; \big\langle\, \theta(f)\, U_a\, g\,\big\rangle_{\rm lattice},
$$
where \(U_a\) is the one-time-step translation of the lattice path integral (Lüscher, CMP 54 (1977) 283, for the pure-gauge sector; extended by Osterwalder–Seiler to include matter fields). \(T\) is self-adjoint on \(\mathcal H_{\rm phys}^{(a,L)}\) because the defining sesquilinear form is reflection-symmetric by Step 2, and it obeys the operator-norm bound
$$
0\;\le\;T\;\le\;1
$$
because \(T\) is built from a Boltzmann weight \(\exp\!\big(-a\times(\text{positive plaquette + matter action density})\big)\) : this is manifestly non-negative (it is the exponential of a real quantity) and bounded above by the identity, since each individual plaquette/hopping factor entering the finite tensor product is itself \(\le1\) in the normalization fixed by the action, and a finite tensor product of operators each \(\le1\) on a compact configuration space remains \(\le1\) . Because \(T\) is bounded, self-adjoint, and satisfies \(0\le T\le1\) , the spectral theorem furnishes a unique self-adjoint logarithm on the orthogonal complement of \(\ker T\) , and one defines the lattice Hamiltonian by
$$
\boxed{\ T \;=\; e^{-aH_a}\,, \qquad H_a \;=\; -\frac1a\log T \;\ge\;0\,, \qquad H_a = H_a^{\dagger}.\ }
$$
The bound \(H_a\ge0\) follows immediately from \(T\le1\) : \(\log T\le0 \Rightarrow -\tfrac1a\log T\ge0\) . Self-adjointness of \(H_a\) follows from self-adjointness of \(T\) together with the spectral calculus applied to the bounded self-adjoint operator \(T\) : \(-\tfrac1a\log(\cdot)\) is a real-valued Borel function on \({\rm spec}(T)\subset[0,1]\) , so functional calculus turns it into another self-adjoint operator. This identity is the operator-theoretic heart of the entire gate: it is the precise statement that the Euclidean lattice construction reconstructs, via the Osterwalder–Schrader correspondence, an ordinary quantum-mechanical time evolution \(e^{-iH_a t}\) (obtained from \(e^{-aH_a}\) by Wick rotation) with a bounded-below, self-adjoint generator acting on a genuinely positive-norm state space — sub-condition (b) of the gate's technical question, established unconditionally at finite \((a,L)\) .

 Step 5 — Perron–Frobenius vacuum and the finite-volume gap. The plaquette-plus-matter Boltzmann weight is strictly positive on the interior of the compact gauge-group configuration space — it is an exponential, hence never zero, and the compact link-integration (Haar) measure has full support. Consequently \(T\) , restricted to the gauge-invariant sector (after the standard removal of the residual finite gauge redundancy at fixed lattice), is an operator with a strictly positive integral kernel in the natural basis. The Perron–Frobenius theorem for strictly positive kernels then gives:
$$
\lambda_0(T) \;=\; \sup\,{\rm spec}(T) \ \text{is a simple eigenvalue, with a strictly positive eigenvector } \Omega,
$$
i.e. the vacuum is unique — no vacuum degeneracy at finite \(a,L\) — and can be represented with strictly positive amplitude in the natural positive basis. Equivalently, \(E_0=-\tfrac1a\log\lambda_0(T)\) is the unique ground-state energy of \(H_a\) , and the finite-volume spectral gap
$$
\Delta(a,L) \;\equiv\; E_1(a,L)-E_0(a,L) \;=\; -\frac1a\log!\Big(\frac{\lambda_1(T)}{\lambda_0(T)}\Big) \;>\;0
$$
is strictly positive at every finite \((a,L)\) , because \(\lambda_1(T)<\lambda_0(T)\) strictly, by Perron–Frobenius simplicity of the top eigenvalue of the connected, aperiodic transfer kernel. In the strong-coupling regime \(\beta<\beta_{\rm conv}\) (the convergence radius of the hopping/cluster expansion around \(\beta=0\) ), the gap additionally admits an analytic lower bound computed order by order in the strong-coupling series (Seiler, LNP 159 , 1982; Münster, 1981), confirming strict positivity of \(\Delta(a,L)\) by an independent, explicit analytic route rather than only the abstract Perron–Frobenius existence argument — with the convergence-radius restriction disclosed rather than silently dropped.

 What Step 5 is not. This finite-volume, finite-cutoff gap is not the Clay Millennium mass-gap statement, and this document does not present it as one. The Clay statement requires the gap to survive uniformly as \(a\to0\) and \(L\to\infty\) together; a finite-lattice gap can, and for a theory that is in fact gapless in the continuum generically does, collapse to zero in that joint limit — the phenomenon the brief names "impostor inflation," a finite-cutoff quantity that is genuinely positive at every finite \((a,L)\) yet is not the surviving continuum invariant. No claim is made or implied that \(\Delta(a,L)\) survives the limit; the claim established in Step 5 is exactly, and only, that \(\Delta(a,L)>0\) for every finite \((a,L)\) — which is precisely what the reflection-positivity-plus-Perron–Frobenius chain in Steps 1–5 delivers, no more and no less.

 III.3 Cross-check: the identity is target-blind and consumes zero dimensionful anchors

 The entire Step 1–5 chain never once invokes any of the four irreducible anchors of the 13-dimensional program, \(M_{\rm Pl}\) , \(\alpha_i(M_Z)\) , \(y_t\) , \(|V_{us}|\) , nor any quantity derived from them such as the compactification scale \(R_0=(2\pi M_U)^{-1}\) , the unification scale \(M_U\) , or the higher-dimensional Planck scale \(M_*\) . The only inputs entering Steps 1–5 are: (i) compactness of the gauge group \(SU(3)\times SU(2)\times U(1)\) at each lattice link — a topological/group-theoretic fact, not a measured number; (ii) positivity of the Boltzmann weight — a sign fact following from the weight being an exponential of a real action density; and (iii) the reflection symmetry \(\theta\mapsto-\theta\) of the lattice action and Haar measure — a discrete-symmetry fact. This is the precise sense in which "no number is borrowed into the finite-cutoff regime": the theorem \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) is a structural , value-free result that would hold verbatim even were every one of the four anchors re-measured tomorrow at different numerical values. It is also the reason UQF-3 can consume only value-free axiomatic inputs — AXIOM-STABILITY-OF-MATTER ( \(H\) self-adjoint, bounded below, ground state exists) and AXIOM-BORN-SIGN (probabilities real and non-negative) — rather than any dimensionful magnitude: this is a positivity/sign theorem, not a magnitude prediction, so there is no \(\sigma\) -band pull against a PDG value anywhere in this construction, and none is reported.

 III.4 The sub-lemma ledger: exactly where the banked chain stops (SL-0 through SL-5)

 The finite-cutoff derivation of §III.2 is the banked bottom half of a six-rung ladder whose top rung is the Clay-class continuum statement. Writing out every rung with its precise status is what prevents the open content from diffusing into an unfalsifiable "positivity is hard" — it isolates the open content to exactly one named analytic statement.

 SL-0 (variational identity). Kernel coercivity \(\Leftrightarrow\) non-existence of a soft (vacuum-orthogonal, energy \(\to0\) ) sequence \(\Leftrightarrow\) spectral gap:
$$
\inf {\rm Spec}(H)\big| {\mathcal H {\rm phys}\ominus{\Omega}} \;>\;0 \quad\Longleftrightarrow\quad \nexists\,{\psi_n}\subset\mathcal H_{\rm phys},\ \psi_n\perp\Omega,\ |\psi_n|=1,\ \langle\psi_n,H\psi_n\rangle\to0 \quad\Longleftrightarrow\quad H\ \text{coercive off}\ \Omega.
$$
This is the min-max/Rayleigh-quotient characterization of a spectral gap for a self-adjoint operator bounded below — a restatement of the spectral theorem, true for any such \(H\) on any Hilbert space, independent of the specific dynamics. It carries no open physical content. BANKED. 

 SL-1 (spectral conversion on \(\mathcal M_4\) ). If the Euclidean two-point function of a local, reflection-positive operator \(\mathcal O\) clusters uniformly,
$$
\big|\langle\mathcal O(x)\mathcal O(0)\rangle - \langle\mathcal O\rangle^2\big| \;\le\; C\,e^{-\Delta|x|}, \qquad |x|\to\infty,
$$
with \(\Delta\) independent of separation, then reflection positivity plus the transfer-matrix representation of the correlator as \(\langle\mathcal O, e^{-|x|H}\mathcal O\rangle\) forces
$$
{\rm Spec}(H)\cap(0,\Delta) = \varnothing, \qquad \Delta \;\ge\; c'\,\Lambda_{\rm YM},
$$
for some order-one structure constant \(c'\) and dynamically generated scale \(\Lambda_{\rm YM}\) . This is a clean conditional implication — clustering \(\Rightarrow\) gap — carrying no fabricated number: \(c'\) and \(\Lambda_{\rm YM}\) remain symbols, exactly as the underlying constructive-QFT literature presents them; the step converts one hypothesis into another, it does not compute either. BANKED (conditional bridge; supplies no number). 

 SL-2 (finite- \(a\) base). The full construction of §III.2: \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) , positive Hilbert space, Perron–Frobenius unique vacuum, strict finite-volume gap \(\Delta(a,L)>0\) , plus the strong-coupling analytic lower bound for \(\beta<\beta_{\rm conv}\) . BANKED / ESTABLISHED — this is the theorem this section derives from first principles.

 SL-3 (continuum kernel construction). Whether the limiting object \(\mathcal H_{\rm phys}^{\rm YM}\equiv\lim_{a\to0,L\to\infty}\mathcal H_{\rm phys}^{(a,L)}\) is itself a genuine, well-defined positive-definite Hilbert space — rather than degenerating, e.g. by the measure concentrating on a lower-dimensional or trivial support, or the limiting generator losing self-adjointness — is the prior , logically antecedent question raised by Gribov (1978), Singer (1978), and Neuberger (1987): even having sidestepped gauge-fixing entirely via Wilson kinematics (which removes the Gribov ambiguity , §III.1), whether a well-defined continuum measure with a positive reproducing kernel exists at all for an interacting, asymptotically-free 4D gauge theory is itself unproven. Bałaban's UV-stability program (CMP, circa 1984–1989) controls the renormalization-group flow of the lattice action toward the continuum, but — stated in the sources without qualification — this is UV stability only ; it does not by itself supply the continuum construction, reflection positivity, or the mass gap. OPEN — the prior question , named and not folded into a later, easier-sounding rung.

 SL-4 (Osterwalder–Schrader axioms H1/H2/H3). Granting SL-3, the next layer is: existence of the continuum Euclidean measure (H1); survival of reflection positivity through the \(a\to0,L\to\infty\) limit rather than being an artifact of the lattice regulator (H2); and nontriviality of the resulting theory — that it is not simply a free (Gaussian) theory in disguise (H3). This is precisely the interacting-sector content of the Clay Millennium Yang–Mills existence-and-mass-gap problem. OPEN — Clay , with no claim made here beyond naming the object correctly.

 SL-5 (Osterwalder–Schrader axiom H4 = Uniform Large-Field Control Bound, "ULFCB"). The single remaining rung — the sharpest possible isolation of what remains unproven — is uniform large-field domination:
$$
\inf_{\substack{\psi\perp\Omega\ |\psi|=1}} \langle\psi,H\psi\rangle \;\ge\; c'\,\Lambda_{\rm YM} \;>\;0 \qquad\text{uniformly through the}\ a\to0,\ L\to\infty\ (\text{OS/IR})\ \text{limit},
$$
equivalently, on the link-variable side, a uniform bound on the field-strength fluctuation,
$$
|\Phi_n|_\kappa \;\le\; K \qquad\text{uniform in}\ n,\,a,\,L,\,\beta,
$$
where \(\Phi_n\) is the accumulated link/plaquette fluctuation at renormalization-group step \(n\) , \(\kappa\) a fixed Hölder-type norm, and \(K\) a constant that must not depend on the cutoff, volume, or bare coupling. This is the wall. No values of \(c'\) , \(K\) , \(\kappa\) satisfying the required uniformity are supplied anywhere in the constructive-QFT literature, and none are supplied here: asserting specific values for them would be fabrication, which this dossier does not commit.

 The mechanism of the obstruction, stated as sharply as the corpus permits. Bałaban's multiscale block-spin renormalization-group analysis shows that large-field configurations — configurations where the plaquette fluctuation departs substantially from its saddle value — are rare : their measure at RG step \(n\) obeys the Gaussian-type suppression
$$
\mu_n(L_n) \;\le\; e^{-c/g(2^na)^2},
$$
with \(g(2^na)\) the running coupling at the effective scale \(2^na\) . This rarity bound is real, proved, and used without modification. But ULFCB does not ask whether large-field configurations are rare (they provably are); it asks whether they are dominated — whether, despite their small measure, they can still host a soft, vacuum-orthogonal sequence of physical states whose energy expectation value is driven to zero. Rarity is a statement about the measure of a set; domination is a statement about whether that set, however small, can host such a sequence. Rarity does not imply domination. Marginal (asymptotically free, non-super-renormalizable) \(d=4\) Yang–Mills is exactly the case where this gap between "rare" and "dominated" carries no spare coercive margin to absorb for free: the theory sits at the borderline where the renormalization-group flow neither generates enough extra suppression to convert rarity into domination automatically, nor rules the possibility out. This is the same difficulty, named without embellishment or euphemism, that has stopped every constructive-QFT program on this exact problem for roughly five decades; SL-5 is not this program's private, bespoke debt — it is the shared debt of the entire field, correctly inherited here rather than manufactured or hidden.

 III.5 The genuinely new structural result: positivity is a proper sub-wall of the mass gap

 The one move in this gate that is not a restatement of published 1970s–1980s constructive field theory is the scope theorem connecting UQF-3 to Gap-02, the Yang–Mills mass-gap problem. Tracing the sub-lemma dependency graph of §III.4 shows the logical arrows run in exactly one direction:
$$
\text{gap-machinery (SL-1: clustering}\Rightarrow\text{gap)} \;\longrightarrow\; \text{positivity refinement (SL-3–SL-5)}, \qquad \text{never the reverse.}
$$
Concretely: SL-2, the finite-cutoff positivity result derived in full in §III.2, requires no input whatsoever from the mass-gap machinery — it holds regardless of whether the continuum theory turns out to be gapped, gapless, or fails to exist as a nontrivial interacting QFT at all. What does require gap-type control is the continuum refinement SL-3–SL-5, and even there the dependency runs one way: gap control (a lower bound on the spectrum away from the vacuum) would supply the coercivity that SL-5's ULFCB demands as a byproduct, but positivity is not a logical prerequisite for the gap. Put in the cleanest possible form: a positive theory need not be gapped (a gapless-but-positive theory — e.g., one with massless excitations — is logically consistent with full reflection positivity), so a hypothetical future proof of continuum reflection positivity (closing SL-3–SL-4, and a weak form of SL-5 sufficient only for boundedness rather than a quantitative gap) would not, by itself, resolve the Clay mass-gap question, whereas a hypothetical future proof of the mass gap (closing SL-1 as an unconditional theorem with an explicit \(\Delta>0\) ) would immediately hand positivity over as a corollary.

 This means the residual UQF-3 carries — achieving SL-3 through SL-5 — is a proper subset of "solve confinement," strictly narrower than the full Gap-02 statement even though the two gates share the identical SL-5/ULFCB technical wall as their respective hardest open ingredient. This is why UQF-3 is tracked as its own gate rather than absorbed into Gap-02, and it is the discipline that forecloses the one inversion error this gate's non-claims explicitly name and forbid: "UQF-3 \(\Rightarrow\) the Yang–Mills mass gap." That inversion is not licensed by anything derived here; the arrow runs gap-machinery \(\to\) positivity, and this document does not run it backward at any point.

 III.6 The exact-rational geometric cross-checks feeding the not-yet-closed \(P(a_6)\) functional

 While the finite-cutoff positivity theorem of §III.2 consumes no geometric magnitude at all, the residual roster names one further, genuinely separate, UQF-3-specific finite object still owed: the order-6 mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient \(a_6^\partial\) on the totally-geodesic fixed locus \(F\) of \(K_6=SU(3)/T^2\) , reached via Levi-Civita transport and Gelfand–Tsetlin (GT) ladder matrix elements. This object is not the Clay wall in disguise — it is finite, in-principle computable, and is honestly carried OPEN/UNMADE with no sign asserted. Its geometric backbone is pinned here to full precision so that a future computation has an unambiguous target and so that the frozen negative controls this program must never contradict are stated inline.

 Exact-rational curvature invariants at the Killing-form normal metric, chamber center \(\vec u=(1,1,1)\) : 
$$
\dim K_6 = 6, \qquad \mathrm{Ric}_i = \frac{5}{12}, \qquad \mathrm{Scal} = \frac{5}{2}, \qquad \mathrm{Scal}^2 = \frac{25}{4},
$$
$$
|\mathrm{Ric}|^2 = \frac{25}{24}, \qquad |\mathrm{Riem}|^2 = \frac{23}{12}, \qquad \frac{|\mathrm{Riem}|^2}{\mathrm{Scal}^2} = \frac{23}{75} = 0.30\overline{6}, \qquad \frac{|\mathrm{Ric}|^2}{\mathrm{Scal}^2} = \frac16, \qquad \frac{\mathrm{Scal}}{\mathrm{Ric}_i} = 6 = \dim K_6.
$$
These five ratios are metric-scale invariant and identical in both the Killing-form normalization used here and the frozen physical \(R_6\) -normalization used elsewhere in the 13D program (in \(R_6\) -norm: \({\rm Ric}_i=1/(2R_6^2)\) , \({\rm Scal}=3/R_6^2\) , and the ratio \({\rm Scal}/{\rm Ric}_i=6\) is unchanged). They are certified against the frozen negative controls named in the geometry pack: \(|\mathrm{Riem}|^2=23/12\) (ratio \(23/75\) ) is never \(31/147\) — a retracted, Bianchi-violating Nomizu-sign-bug artifact — and \(|\mathrm{Riem}|^2\) is never \(60\) , the distinct value belonging to the unrelated manifold \(S^6\) in its round-unit normalization. The companion cubic (weight-6) invariants at the same Einstein center are \(K_1 = R_{ab}{}^{cd}R_{cd}{}^{ef}R_{ef}{}^{ab} = -113/72\) and \(K_2 = R_{abcd}R_{aecf}R_{ebfd} = -5/72\) , with \(|\nabla{\rm Riem}|^2 = 1/4 \ne 0\) — the fact certifying that \(K_6\) is homogeneous but not locally symmetric, which is precisely why the graviton leg of the \(a_6\) heat-kernel calculation carries a nontrivial Gelfand–Tsetlin off-diagonal ladder term rather than vanishing identically as it would on a symmetric space.

 The two independent heat-kernel calibration routes, agreeing to \(\sim4\times10^{-14}\) . Before the machinery is pointed at the unknown \(K_6\) graviton sector, it is calibrated on round spheres of known exact spectrum:
$$
a_6(S^2) = \frac{4}{315}, \qquad a_6(S^4) = \frac{74}{63}, \qquad a_6(S^6) = \frac{1139}{63}, \qquad a_6^{\rm conformal}(S^6) = \frac{5}{63}.
$$
The \(S^6\) round-unit row additionally calibrates the \(a_4\) formula independently, returning exactly \(12\) — a passed control confirming \(K_6\) is not being mistaken for \(S^6\) anywhere in the pipeline. These four values are not themselves the target \(a_6^\partial\) ; they are the control set whose two-route agreement at \(\sim10^{-14}\) precision licenses confidence in the same machinery once it is applied to \(K_6\) .

 The scale-free (Levi-Civita-normalization-immune) cross-check: 
$$
\frac{a_4}{a_2^{\,2}} = \frac{66}{125} = 0.528,
$$
robust to the choice of overall curvature normalization by construction, since both numerator and denominator scale identically under a constant rescaling of the metric.

 The \(\mathbb Z_2\) orbifold-defect order-6 product-trace check , exercising exactly the fixed-locus structure that \(a_6^\partial\) will eventually need: on \(S^2\times(S^1/\mathbb Z_2)\) , the exact heat-kernel product rule \(a_{2k}(M_1\times M_2)=\sum_{i+j=k}a_{2i}(M_1)a_{2j}(M_2)\) gives
$$
\tfrac12\cdot\frac{4}{315} \;=\; \frac{2}{315},
$$
peeled two independent ways to agreement at \(\sim1\times10^{-5}\) (direct route; this level of residual noise is disclosed as "peel-noise," not hidden) and \(\sim1.3\times10^{-14}\) (an independent Vandermonde route). Separately, the twisted-circle trace
$$
\mathrm{Tr}\big(\sigma\,e^{-tD}\big) \;=\; 1.000000000000\ldots,
$$
is exactly \(t\) -independent, built from integer powers only — there is no half-integer boundary tower . This flat, integer-only result is the structural fact that dissolves a tempting but wrong prior framing in which "the boundary \(a_6\) " would show up as an infinite tower of fractional-power corrections requiring an unbounded computation; the equivariant structure instead collapses to a single, finite, well-posed heat-kernel coefficient, which is exactly why \(a_6^\partial\) is a nameable, in-principle-finite computation debt rather than an open-ended analytic wall of the SL-5 type.

 The \(\mathbb Z_2\) Donnelly equivariant defect structure — the piece that is fully built, not owed. Two isolated fixed points on \(S^1_Y/\mathbb Z_2\) at \(\theta=0,\pi\) ; reflection \(g\) -trace \(\sum_{\rm fixed} 1/|1-dg| = 2\times\tfrac12 = 1\) ; per-fixed-point \(a_0\) defect \(+1/4\) for even-parity fields and \(-1/4\) for odd-parity fields; \(\det(I-d\sigma|_N)=2\) at each fixed point (the reflection acts as \(-1\) on the normal 2-plane); the fixed locus \(F\) is totally geodesic with zero angle deficit; and \(\mathrm{Vol}(S^1_Y/\mathbb Z_2) = \pi R_Y = 1/(2M_U)\) at the chamber center. These combine into the exact Donnelly equivariant trace identity
$$
\mathrm{tr}[a_6]^{\mathbb Z_2} \;=\; \tfrac12\,c_3^\gamma,
$$
the precise structural relation that will convert a future value of the constant \(c_3^\gamma\) into the boundary object \(a_6^\partial\) and thence into the candidate positivity functional \(P(a_6)\) . The structure is complete and exact; the numeric value of \(c_3^\gamma\) , and hence of \(a_6^\partial\) and \(P(a_6)\) , remains OPEN and UNMADE — no sign is asserted anywhere in this dossier, and none is fabricated to force a closure. A definite-sign violation of \(P(a_6)\ge0\) , were it ever computed, would itself constitute a legitimate refuting close of this residual: a decidable negative outcome is an honest terminal for this specific sub-object, not a failure of the broader program.

 Grading constants forced by linear algebra on \(D=13\) plus the reflection structure — carried exactly, three distinct roles, never conflated. With the fibre involution \(A=\mathrm{diag}(1_{12},-1)\) on the 13-dimensional fibre (the \(\mathbb Z_2\) reflection generator restricted to the fibre):
$$
\mathrm{tr}\,A = 11, \qquad \mathrm{tr}\,A^2 = 13,
$$
$$
\mathrm{tr}\,\mathrm{Sym}^2(A) \;=\; \frac{\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2}{2} \;=\; \frac{13+121}{2} \;=\; \frac{134}{2} \;=\; 67.
$$
This graded trace is identified with the graded graviton weight, \(\mathrm{tr}\,\gamma_{\rm grav}=67\) ; the graded ghost weight is \(\mathrm{tr}\,\gamma_{\rm ghost}=\mathrm{tr}\,A=11\) ; and the Block-A graded weight is
$$
67 - 2(11) \;=\; 45,
$$
a combination forced by \(D=13\) together with the reflection involution, not an independent input. Firewall, stated explicitly because the two numbers are easy to conflate: the graded graviton weight \(67\) is a categorically different quantity from the ungraded \(\mathrm{Sym}^2\) multiplet count on the same 13-dimensional fibre, \(\dim\,\mathrm{Sym}^2(\mathbb R^{13}) = \binom{14}{2} = 91\) . The two numbers answer two different questions — a graded trace under the reflection involution \(A\) versus a total ungraded multiplet dimension — and both are correct in their own role; this document never substitutes one for the other, and the exact identity \((13+121)/2=67\) is the internal cross-check that pins the graded value independently of the ungraded count.

 III.7 Why this is the correct terminal: CERTIFIED-IRREDUCIBLE, stated exactly

 Collecting §III.2 through §III.6: the finite-cutoff transfer-matrix identity \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) , together with Perron–Frobenius vacuum uniqueness and the strict finite-volume gap \(\Delta(a,L)>0\) , is a complete, unconditional, target-blind derivation resting on published theorems — Osterwalder–Schrader (1973, 1975), Osterwalder–Seiler (1978), Lüscher (1977), Seiler (1982), Münster (1981) — applied without modification of their hypotheses to the frozen 13-dimensional branch's gauge-invariant Wilson kinematics. This is SL-2, banked in full in §III.2. The sub-lemma rungs beneath and around it, SL-0 and SL-1, are likewise banked: one as pure spectral theory carrying no open content, the other as a clean conditional bridge that derives no fabricated numerical value. The single blocking rung, SL-5 (ULFCB, equivalently Osterwalder–Schrader axiom H4, required uniformly through the continuum limit), is exactly and only the interacting-sector content of the Clay Millennium Yang–Mills existence-and-mass-gap problem — an external wall the entire constructive-QFT field has faced for roughly five decades, not a bespoke shortfall of this thirteen-dimensional construction. It is inherited here, named with its precise mechanism (Bałaban rarity, established and used; domination, not established and not fabricated), and dispositioned REDUCED-TO-AXIOM under the granularity screen P1 — the finite-cutoff theory is declared the physically relevant object, so the never-completed \(a\to0,L\to\infty\) limit is dissolved onto that postulate rather than left as a bare, unresolved conjecture, a dissolution onto a named axiom that is explicitly not the same act as solving the underlying analytic problem and is not presented as such anywhere in this derivation.

 Separately from that shared wall, the one finite object that is this gate's own — the \(a_6^\partial\) boundary heat-kernel coefficient of §III.6 — is carried honestly as OPEN/UNMADE, with its complete equivariant scaffolding in place and no sign guessed. And the one genuinely new logical contribution of this gate, the scope theorem of §III.5, narrows UQF-3's residual to a proper subset of the harder Gap-02 statement rather than allowing it to inherit the full mass-gap problem by default. Because the banked half of the ladder is a real, unconditional, non-trivial theorem; because the blocking half is precisely named rather than left diffuse; because no fabricated constant closes SL-5 and no guessed sign closes \(P(a_6)\) ; and because the scope theorem correctly bounds how much of the Clay problem this gate actually owes — the terminal CERTIFIED-IRREDUCIBLE / RESOLVED +0 is the exact, non-inflated, non-deflated description of where this construction stands: a genuine theorem banked cleanly on one side of a correctly identified, correctly shared, currently uncrossable wall.

 The insights that made it work

 Insight 1 — attack the object the granularity axiom actually licenses, not the object the Clay problem demands. The single move that converts UQF-3 from "another unsolved Yang–Mills positivity problem" into a gate with a genuine, unconditional, banked result is a re-targeting of the claim, not a weakening of it. The program's granularity postulate P1 — the axiom, load-bearing everywhere in the frozen 13-dimensional theory, that the physically relevant object is the theory at finite lattice spacing \(a>0\) and finite volume \(L<\infty\) , not the idealized \(a\to0\) , \(L\to\infty\) continuum — is not invoked here as a rhetorical hedge. It is invoked as a proof strategy . Once the physically-relevant theory is fixed to be the finite-cutoff theory, reflection positivity stops being an open constructive-QFT conjecture and becomes an application of three theorems that have been sitting in the literature, fully proved, for nearly fifty years: Osterwalder–Seiler lattice reflection positivity (1978), Lüscher's transfer-matrix construction (1977), and the strong-coupling spectral gap of Seiler (1982) and Münster (1981). The insight is a category change : the gate does not need to invent new mathematics to win a real, unconditional result — it needs to correctly identify which mathematics already proves the theorem for the object the program's own axioms say is the physical one. This is the same move that appears throughout the framework program under the name "granularity ⇒ MDL": once P1 fixes finite \(a\) as physical, the minimum-description-length object is the finite transfer matrix, and every downstream positivity statement is judged against that object, not against an idealized limit the theory never claims to need. Reflection positivity at finite cutoff is therefore not a partial result awaiting completion — it is the complete and correct statement of the physical claim, full stop, on the frozen branch.

 Why this is not evasion. The reason this re-targeting is honest rather than a target-loading trick is that it does not touch the hard part of the problem — it isolates it. Every legitimate open question in constructive Yang–Mills theory survives the re-targeting exactly where it lives: in the uniformity of the limit \(a\to0\) , not in the finite- \(a\) statement. The dossier is explicit that the two objects are logically and mathematically distinct: a finite-lattice spectral gap \(\Delta(a,L)>0\) can, and generically does, collapse as \(a\to0\) ("impostor inflation" — a finite-volume/finite-cutoff positive quantity that is not the surviving Clay-class statement). Recognizing this distinction crisply is itself part of the insight: it lets the certificate claim exactly what it has proved (finite- \(a\) positivity, unconditionally) and refuse exactly what it has not proved (uniform continuum positivity), with a bright line between the two rather than a blurred continuum of partial credit.

 Insight 2 — Wilson kinematics dissolve the Gribov ambiguity before it can appear. The second load-bearing insight is a choice of representation for the finite-cutoff theory: gauge-invariant Wilson lattice variables (link variables in the compact gauge group, plaquette actions), not a gauge-fixed continuum action. This choice is not incidental — it is what allows the finite-cutoff reflection-positivity leg to be unconditional rather than conditional on a gauge-fixing prescription. The Gribov–Singer obstruction (Gribov 1978; Singer 1978; sharpened by Neuberger 1987) is the standard no-go result that any continuum gauge-fixing of a non-abelian gauge theory admits Gribov copies — the gauge slice is not global, so a naive Faddeev–Popov construction of a positive kernel on gauge orbit space fails globally. This obstruction is exactly what makes SL-3 (continuum kernel construction) and the retained-sector BRST leg conditional and delicate. But on the Wilson lattice with unfixed gauge , there is no gauge-fixing step at all: the path integral is over compact link variables with a manifestly positive, gauge-invariant Boltzmann weight, and reflection positivity is a property of that weight directly — it does not route through a gauge slice, and therefore the Gribov ambiguity has no purchase on it. This is why SL-2 (the finite- \(a\) base construction, §3.1 in the derivation chain) is banked as ESTABLISHED/DERIVED without qualification, while the BRST no-ghost leg (§3.2, which does require gauge-fixing to define \(Q_{\rm BRST}\) and the Kugo–Ojima quartet mechanism) is graded only CERTIFICATE-CONDITIONAL. The insight is precise: which representation you positivity-certify in matters as much as what you are trying to prove , and choosing the gauge-invariant representation for the piece of the argument that can be made unconditional is what banks SL-2 cleanly, while honestly leaving the gauge-fixed BRST piece exposed to its own genuine dependency (on UQF-4's \(Q^2=0\) ).

 Insight 3 — a completeness argument, not a computation, produces the transfer-matrix positivity structure. The chain from "reflection positivity of the Euclidean measure" to "positive physical Hilbert space with self-adjoint bounded Hamiltonian" is not a numerical estimate; it is the Osterwalder–Schrader reconstruction machinery applied to a well-defined finite object. Concretely: reflection positivity of the lattice measure with respect to time-slice reflection \(\theta\) means \(\langle \theta(f), f\rangle \ge 0\) for observables \(f\) supported on the positive-time half-lattice. This single positivity property is equivalent , via the GNS-type reconstruction that Osterwalder–Seiler carry out explicitly for lattice gauge theory, to the existence of a Hilbert space \(\mathcal H_{\rm phys}^{(a,L)}\) on which a self-adjoint transfer operator \(T\) with \(0\le T\le 1\) acts, and setting \(T\equiv e^{-aH_a}\) defines a Hamiltonian \(H_a\ge0\) automatically — boundedness below is not an extra assumption, it falls out of \(T\le1\) for free, because \(T=e^{-aH_a}\le1 \iff H_a\ge0\) for \(a>0\) . This is the structural reason the "no-ghost" conclusion at finite cutoff requires no separate spectral analysis: positivity of the reflection form is positivity of the norm on the reconstructed space, by the same equivalence that Osterwalder–Schrader established for the continuum reconstruction theorem, specialized here to the finite lattice where Osterwalder–Seiler proved it holds. Layered onto this, the Perron–Frobenius theorem (applicable because \(T\) is a positive operator, in the order sense, on the finite-dimensional or trace-class Hilbert space at finite \(a,L\) ) gives a simple (non-degenerate) vacuum eigenvalue — this is where "unique ground state" comes from, not from an added axiom. And the strong-coupling expansion (valid for bare coupling \(\beta<\beta_{\rm conv}\) , a genuine convergence-radius restriction, not a universal claim) supplies an explicit, analytically controlled lower bound on the spectral gap \(\Delta(a,L)\) above the vacuum — so "gap exists" is not asserted but shown , in the regime where the expansion converges. Every one of these three sub-results — positive Hilbert space, simple vacuum, finite-volume gap — is a structural consequence of the single reflection-positivity input, not three independent claims requiring three independent proofs. That structural economy is itself an insight: one positivity property, correctly established at finite cutoff, propagates through standard functional-analytic machinery to deliver the entire finite-theory package for free.

 Insight 4 — the SL-0…SL-5 sub-lemma decomposition turns a diffuse "open problem" into one precisely named analytic statement, which is what makes the wall nameable rather than hand-wavy. A generic complaint against a constructive Yang–Mills-type claim is that "positivity in the continuum is open" is too diffuse to certify anything against — it invites either false despair (everything is open) or false optimism (surely some clever trick closes it). The actual insight embedded in this gate's derivation chain is a decomposition of the continuum argument into six sub-lemmas, each independently gradable:
- SL-0 (variational identity: kernel coercivity \(\Leftrightarrow\) no soft sequence \(\Leftrightarrow\) spectral gap) is a piece of pure spectral theory — it carries no open physical content because it is an equivalence between three characterizations of the same abstract fact, true for any self-adjoint operator bounded below with discrete-or-continuous spectrum, independent of the dynamics.
- SL-1 (the 4D spectral-conversion statement, that uniform exponential clustering of correlators implies a genuine spectral gap \(\Delta \ge c'\Lambda_{\rm YM}\) ) is banked as a clean conditional : if uniform clustering holds, then the gap follows, by standard operator theory — it derives no numerical value and asserts no unproven input; it is a correct implication, not a completed existence proof.
- SL-2 is the finite- \(a\) base (already discussed) — fully banked.
- SL-3 (does the limiting object \(\mathcal H_{\rm phys}^{\rm YM}\) even exist as a genuine positive Hilbert space in the continuum) is where the Gribov/Singer/Neuberger obstruction lives — this is the prior question , logically upstream of positivity itself: you cannot ask whether the continuum inner product is positive until you know the continuum construction exists at all.
- SL-4 (Wightman/Osterwalder–Schrader axioms H1–H3: does a continuum measure exist, does reflection positivity survive the limit, is the theory nontrivial) is exactly the Clay Millennium formulation.
- SL-5 (axiom H4, "uniform large-field bound," equivalently \(\inf\langle\psi,H\psi\rangle \ge c'\Lambda_{\rm YM}>0\) uniformly through the limit, equivalently a uniform bound \(\|\Phi_n\|_\kappa \le K\) on the link measure independent of \(n,a,L,\beta\) ) is identified as the irreducible core — the wall inside the wall.

 The insight is that this decomposition is not a stalling tactic; it is diagnostic. It lets the dossier state, with full honesty, that four of six sub-lemmas (SL-0, SL-1 conditionally, SL-2 unconditionally) are banked, and that the open content of "continuum positivity" collapses to exactly one analytic statement (SL-5/ULFCB) rather than remaining an amorphous "it's all still open." Even the character of why SL-5 resists is stated precisely, and this is itself an insight worth banking: the large-field region of the lattice configuration space is provably rare — Bałaban's estimate \(\mu_n(L_n) \le e^{-c/g(2^na)^2}\) shows the measure of large-field configurations is Gaussian-small in the renormalized coupling — but rarity is not domination . ULFCB does not ask "how much measure sits on large fields" (answered: very little); it asks whether the rare large-field directions can nonetheless host a soft, vacuum-orthogonal physical excitation that survives the \(a\to0\) limit — a question about the coercivity margin of the kernel on those rare directions, not about their measure. Marginal \(d=4\) Yang–Mills (the critical/log-divergent dimension for the relevant renormalization-group flow) is identified as leaving no spare coercive margin to answer this on general grounds — which is precisely the same razor's-edge structural feature that makes the interacting-sector Clay problem hard for every research program, not a defect specific to this one. Naming the mechanism of the obstruction (rarity vs. domination) rather than just its label ("open") is what makes the wall a characterized wall — inherited and precisely described — rather than a black box labeled "TODO."

 Insight 5 — a genuinely new scope result: positivity is a proper sub-wall of the mass gap, and the dependency arrow only runs one way. Beyond correctly locating and decomposing the known wall, this gate produces one result that is not simply "citing Osterwalder–Seiler more carefully" — a scope theorem with real content: UQF-3's positivity problem is a proper sub-wall of the Gap-02 mass-gap problem; positivity does not require the mass gap. The logical content is that the gap-machinery (SL-1's clustering \(\Rightarrow\) spectral-gap argument, and the finite-volume gap of SL-2) feeds into the positivity/domination question, never the reverse: you can ask "is \(H_a\ge0\) with the right spectral structure" as a question that is answered by properties of the transfer matrix and the reflection form directly (SL-0 through SL-2), without first assuming a mass gap exists in the continuum. This means the residual UQF-3 carries is strictly narrower than "prove confinement" or "solve the mass-gap problem" — a positivity certificate could in principle be established (or refuted) by resolving SL-5 alone, without independently having to establish the existence of a mass gap as a separate prerequisite. Recognizing and stating this one-way dependency is what prevents a subtle and easy-to-make error: inverting the arrow into "UQF-3 \(\Rightarrow\) Yang–Mills mass gap," which is explicitly and correctly forbidden in this gate's non-claims. The insight is a discipline about direction of implication in a coupled system of open problems: knowing which of two related unsolved statements is upstream of the other is itself nontrivial content, and it is what lets this dossier scope its residual precisely rather than inheriting the full mass-gap problem wholesale.

 Insight 6 — why the boundary/orbifold geometry is the natural home for the one UQF-3-specific finite open object, and why it is not the Clay wall in disguise. Separately from the shared continuum wall, the gate carries one finite, in-principle-computable open object of its own: the order-6 mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient \(a_6^\partial\) on the totally-geodesic fixed locus \(F\) of \(K_6=SU(3)/T^2\) , computed via Levi-Civita transport and Gelfand–Tsetlin ladder combinatorics. The reason this object exists and is the right one to name is geometric, not accidental: the frozen 13-dimensional arena's boundary structure is the orbifold \(S^1_Y/\mathbb Z_2\) , whose \(\mathbb Z_2\) reflection \(\theta\mapsto-\theta\) has exactly two isolated fixed points at \(\theta=0,\pi\) . Reflection positivity is, physically, a statement about reflecting across a hypersurface — and on this frozen branch the only built-in reflection is precisely this orbifold action, which is why its fixed-point data is where a would-be finite positivity functional \(P(a_6)\ge0\) must be built from. The Donnelly equivariant-trace formalism gives the exact combinatorial content of this fixed-point structure without any free parameter: two isolated fixed points, each contributing \(1/|1-dg| = 1/|1-(-1)| = 1/2\) to the equivariant trace, for a total \(\sum 1/|1-dg| = 2\times\tfrac12=1\) ; the per-fixed-point \(a_0\) -defect is \(+1/4\) for even parity and \(-1/4\) for odd parity, and \(\mathrm{tr}[a_6]^{\mathbb Z_2} = \tfrac12 c_3^\gamma\) is the exact equivariant relation linking the still-unknown \(a_6^\partial\) boundary coefficient to the constant \(c_3^\gamma\) that would feed the positivity functional. This is a structural insight, not a numerical one: it explains why the orbifold fixed-point locus, and not some other feature of the 13-dimensional geometry, is the correct and only candidate site for a UQF-3-specific finite positivity object, and it is why this object is honestly kept OPEN/UNMADE (no sign asserted, no value fabricated) rather than either assumed to vanish or invented. The insight that dissolves a wrong framing here is worth stating explicitly: computing the twisted-circle trace \(\mathrm{Tr}(\sigma e^{-tD})\) on this orbifold returns the exact constant \(1.000000000000\ldots\) , \(t\) -independent, built from integer powers only — there is no half-integer boundary tower in this trace. That flat, integer-only result is what dissolves a tempting but wrong picture in which the "boundary \(a_6\) " object would show up as some infinite tower of fractional-power boundary corrections requiring an unbounded computation; instead the equivariant structure collapses to a single finite well-posed heat-kernel coefficient, which is exactly why \(a_6^\partial\) is a nameable, in-principle-finite computation debt (owed, at the Gelfand–Tsetlin ladder-matrix-element stratum) rather than an open-ended analytic wall of the SL-5 type. Recognizing that these are two different species of "open" — a finite uncomputed coefficient (R4/ \(a_6^\partial\) ) versus an infinite-dimensional constructive limit (R3/SL-5) — and refusing to let either one borrow the other's difficulty class, is itself part of what makes the certificate honest: R4 is not quietly promoted to "as hard as the Clay problem," and R3 is not quietly demoted to "just a finite computation we haven't gotten to yet."

 Insight 7 — why the linearized-graviton and free-field results are legitimate certificates rather than smuggled continuum claims. The dossier bank two further positive-norm results — the two transverse-traceless linearized graviton polarizations (van Nieuwenhuizen 1973; Stelle 1978) and the free-field equivalence between the Osterwalder–Schrader Euclidean picture and BRST cohomology — and the insight in each case is recognizing the precise boundary of validity that keeps the result honest. The linearized graviton result is DERIVED-GIVEN-E: it is a statement about the free/linearized theory at the level of the frozen endomorphism data (the Lichnerowicz operator \(E_L\) on \(\mathrm{Sym}^2_0\) , with certified trace data \(\mathrm{tr}\,E_L = 40/3\) , \(\mathrm{tr}\,E_L^2 = 241/18\) at the Killing-form center on the 20-dimensional transverse-traceless bundle) — it is unconditionally true at that level and is not extended, by this gate, to claim anything about the interacting theory above the KK cutoff, which is explicitly carried forward as an open export (R7, to UQF-14). Likewise the OS \(\leftrightarrow\) BRST equivalence is standard, textbook material for free fields ; the insight is refusing to let a free-field textbook citation stand in for a proof on the interacting, frozen branch — interacting-theory OS \(\leftrightarrow\) BRST equivalence folds back into the same R3/R6 continuum-construction residual rather than being treated as separately established. In both cases the discipline is the same one running through the whole gate: identify exactly the regime (linearized; free) in which a classical, textbook result is unconditionally true, quote it there without embellishment, and refuse to let the regime boundary silently dissolve when the result is carried into the write-up.

 Why the whole assembly is reproducible and not merely asserted. Every quantitative or structural ingredient above traces to a named, dated, peer-reviewed result (Osterwalder–Schrader 1973/1975; Osterwalder–Seiler 1978; Lüscher 1977; Seiler 1982; Münster 1981; Kugo–Ojima 1979; Gribov 1978; Singer 1978; Neuberger 1987; Bałaban c. 1984–89; van Nieuwenhuizen 1973; Stelle 1978) applied, without modification of its hypotheses, to the specific frozen 13-dimensional object \(\mathfrak B_{\rm active} = [\mathcal M_4\times K_6\times S^2\times S^1_Y]\oplus[\mathcal F^+_{\rm finite}\oplus\mathcal C_{\rm admiss}]\otimes[\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\oplus\mathcal E_{\rm proton}]\) with \(K_6=SU(3)/T^2\) and \(D=13\) . No new axiom is introduced beyond the granularity postulate P1 (already load-bearing across the whole program) and the two value-free physical requirements the gate consumes — AXIOM-STABILITY-OF-MATTER (the Hamiltonian is self-adjoint, bounded below, with a ground state) and AXIOM-BORN-SIGN (probabilities are real and non-negative) — neither of which is a dimensionful magnitude fitted to data, which is precisely why this gate reports no \(\sigma\) -pulls: there is nothing here to tune against a measured number, only exact-rational geometric cross-checks (the Killing-form curvature invariants \(\mathrm{Ric}_i=5/12\) , \(\mathrm{Scal}=5/2\) , \(|\mathrm{Riem}|^2=23/12\) with ratio \(23/75\) , and the topological Euler characteristics \(\chi(K_6)=6\) , \(\chi(S^2)=2\) , \(\chi(S^1_Y/\mathbb Z_2)=1\) ) that any independent reproduction of the orbifold fixed-point / heat-kernel structure must land on exactly, with no free parameter available to fudge a mismatch.

 The synthesis. Stripped to its logical skeleton, the reasoning that makes UQF-3 believable is: (1) fix the claim to the object the program's own granularity axiom says is physical (finite \(a\) , finite \(L\) ) rather than to the idealized continuum the field at large has never constructed; (2) choose the gauge-invariant Wilson representation for that finite object so the Gribov obstruction — a genuine, proven no-go for gauge-fixed continuum constructions — has no purchase on the finite-cutoff leg; (3) let the Osterwalder–Schrader/Osterwalder–Seiler reconstruction equivalence do the work of turning one positivity property of the reflection form into the entire package of positive Hilbert space, self-adjoint bounded-below Hamiltonian, simple vacuum, and finite gap, since these are not four separate claims but one structural consequence viewed four ways; (4) decompose the remaining continuum question into six sub-lemmas so that "open" collapses from a diffuse verdict into one precisely named analytic statement (SL-5/ULFCB) whose mechanism of resistance (rarity without domination, at the coercivity-starved marginal dimension \(d=4\) ) is itself understood and stated, not just labeled; (5) prove a genuine scope theorem — positivity is a sub-wall of the mass gap, never the reverse — that narrows the residual rather than inheriting the whole Clay problem by default; (6) locate the one UQF-3-specific finite open object at the only geometrically privileged reflection locus the frozen branch actually has (the \(S^1_Y/\mathbb Z_2\) fixed points), using the exact Donnelly equivariant-trace combinatorics to show it is a nameable finite computation debt, not an open-ended tower; and (7) keep every linearized/free-field result honestly scoped to the regime in which it is textbook-true, exporting its extension to the interacting or above-cutoff regime as a named residual rather than a silent claim. None of these seven moves manufactures a new theorem the field does not already possess; the achievement is entirely in the targeting, decomposition, and honest boundary-drawing — which is exactly the kind of insight that is reproducible by any reader who checks the same published theorems against the same frozen geometric object.

 Evidence & reproducibility

 0. What this gate can and cannot be checked against

 UQF-3 is graded CERTIFIED-IRREDUCIBLE / RESOLVED +0 , and that grade fixes what "evidence" means here before a single number is quoted. This gate consumes zero of the four irreducible dimensionful anchors \(\{M_{\rm Pl},\ \alpha_i(M_Z),\ y_t,\ |V_{us}|\}\) — it is scale-blind by construction, present in the derivation only through the shared frozen radii of the 13-dimensional arena, never as a tuned magnitude. Its two consumed inputs, AXIOM-STABILITY-OF-MATTER (the Hamiltonian is self-adjoint and bounded below, with a genuine ground state) and AXIOM-BORN-SIGN (probabilities are real and non-negative), are value-free structural axioms, not numbers with error bars. Consequently there is no PDG band to pull against and no \(\sigma\) to report for this gate's central claim — reporting a fabricated pull here would itself be exactly the fabrication error this discipline exists to prevent.

 What a reader can independently verify falls into three kinds, each treated in full below: (1) the finite-cutoff reflection-positivity claim itself, which is a structural pass/fail against the hypotheses of three named, dated, peer-reviewed theorems, not a numerical fit; (2) a web of exact-rational geometric identities on \(K_6 = SU(3)/T^2\) that the still-open finite object ( \(a_6^\partial\) , the mixed-boundary heat-kernel coefficient) must be consistent with, every one of them re-derivable by hand from the root system with no external input; and (3) a set of frozen negative controls — specific wrong values the corpus has identified, retracted, and keeps live as tripwires — such that a correct independent re-derivation must land clear of every one of them. A full from-scratch reproduction recipe closes the section.

 1. The structural pass/fail checks (no \(\sigma\) -pulls — none fabricated)

 Check 1 — finite-cutoff transfer-matrix positivity (DERIVED, unconditional). The claim under test: at fixed lattice spacing \(a>0\) and finite volume \(L<\infty\) , using gauge-invariant Wilson-loop kinematics (compact-group link variables, plaquette action, no gauge-fixing), the Euclidean transfer operator
$$
T = e^{-aH_a}, \qquad 0 \le T \le 1, \qquad H_a \ge 0,
$$
is self-adjoint, has spectrum confined to \([0,1]\) , possesses a simple (non-degenerate) Perron–Frobenius vacuum, and a strictly positive finite-volume spectral gap \(\Delta(a,L)>0\) . This is not fitted to any measured quantity; it is the content of Osterwalder–Seiler reflection positivity for lattice gauge theory (Ann. Phys. 110 (1978) 440) combined with Lüscher's explicit transfer-matrix construction (Comm. Math. Phys. 54 (1977) 283). The pass/fail test is whether the hypotheses of those theorems are met by the frozen branch's gauge content: a compact gauge group ( \(SU(3)_c\times SU(2)_L\times U(1)_Y\) , always compact at finite cutoff — no non-compact link variable is ever introduced anywhere in the frozen 13-dimensional construction), a reflection-positive plaquette action under time-slice reflection, and Haar-measure positivity on the link variables. Every one of these hypotheses is satisfied by inspection on the frozen branch, because the branch introduces no modification to the standard Wilson kinematics that the cited theorems were proved for. Verdict: PASS, unconditionally, at any finite \((a,L)\) . There is no free parameter in this check and no place a number could be fudged — either the gauge group is compact and the action is the standard plaquette action (it is), or it is not.

 Check 2 — strong-coupling analytic gap (DERIVED/ESTABLISHED, scope-limited). For bare coupling \(\beta<\beta_{\rm conv}\) (the convergence radius of the character/cluster expansion), the expansions of Seiler (LNP 159 , 1982) and Münster (1981) prove \(\Delta(a,L)>0\) analytically rather than merely asserting it non-constructively. This, too, is a structural theorem with a named, checkable domain of validity, not a fitted magnitude. The honest scope caveat, carried forward without softening: strong coupling sits on the opposite side of the continuum crossover from the regime an asymptotically free theory needs at \(a\to0\) , so this finite- \((a,L)\) , finite- \(\beta\) gap is explicitly not the Clay-class continuum gap, and no claim to the contrary is made anywhere in this dossier. A finite-lattice gap of this kind can collapse as \(a\to0\) — the corpus names this failure mode explicitly as "impostor inflation" and refuses to commit it. Verdict: PASS for \(\beta<\beta_{\rm conv}\) ; explicitly not extrapolated to \(\beta\to\infty\) . 

 Check 3 — perturbative BRST no-ghost (CERTIFICATE-CONDITIONAL). On the retained 4D zero-mode + low-KK sector, Kugo–Ojima (1979) quartet cancellation — checked order by order in the loop expansion on the retained field content \(\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\) — removes the longitudinal and time-like gluon polarizations against the ghost–antighost pair in every \(S\) -matrix element checked, by the standard alternating-sign BRST-doublet counting argument. This check passes at every loop order examined, but its grade is honestly capped at CERTIFICATE-CONDITIONAL because it inherits BRST nilpotency \(Q^2=0\) as an audited , not re-derived, input, with the anomaly-closure content of that audit pointed to UQF-4 rather than re-proved here. Matter and Higgs sectors are positive-norm by construction (no BRST subtlety); the linearized graviton carries exactly its 2 transverse-traceless positive-norm polarizations (DERIVED-GIVEN-E, linearized only — see §2 below for the exact trace data).

 The scope theorem as a checkable claim, not just an assertion. One further "check" belongs in this list because it is a genuinely new, checkable logical statement rather than a citation: the derivation chain shows explicitly that the dependency between UQF-3 and the Gap-02 mass-gap problem runs strictly one way, gap-machinery → positivity , never the reverse. Concretely, SL-0 through SL-2 (the variational spectral identity, the finite-volume gap, and the finite-cutoff transfer-matrix construction) can each be verified as stated without any prior assumption that a continuum mass gap exists; nothing in Checks 1–3 above invokes confinement or a mass gap as a hypothesis. A reader auditing this dossier for a hidden circularity — "does UQF-3 secretly assume the very mass-gap result Gap-02 is trying to prove?" — can verify directly that it does not: every theorem cited in Checks 1–3 has hypotheses statable and checkable purely in terms of reflection positivity of the lattice measure and compactness of the gauge group, with no mass-gap premise anywhere in the hypothesis list.

 Bottom line for this subsection. Zero entries belong in a pull table for UQF-3, and zero are reported. What a reader gets instead is three theorem-hypothesis checks (PASS / PASS / PASS-conditional) plus one scope theorem verifiable by inspecting that no cited theorem's hypothesis list smuggles in a mass-gap premise.

 2. Internal consistency cross-checks (exact-rational, target-blind)

 Although UQF-3's central claim carries no dimensionful number, the one UQF-3-specific finite object still open — the order-6 mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient \(a_6^\partial\) on the totally geodesic fixed locus \(F\) of \(K_6=SU(3)/T^2\) , which would feed a coefficient \(c_3^\gamma\) into a would-be decision functional \(P(a_6)\ge0\) — sits inside a lattice of exact-rational geometric invariants that are independently computable today, to extremely high precision, with no free parameter. These cross-checks matter specifically to this gate because they are the negative-control backbone any eventual \(a_6^\partial\) computation must be consistent with: an internal contradiction anywhere in this lattice would be a red flag for the entire boundary-object program, not a cosmetic slip.

 (a) Curvature invariants at the Killing-form center, reproduced from the root system alone. The \(K_6=SU(3)/T^2\) flag manifold's invariant metric at the symmetric chamber center \(\vec u=(1,1,1)\) gives, in the Killing-form normal-metric normalization ( \(g=(-B)|_{\mathfrak m}\) , \(B(X,Y)=6\,{\rm Tr}(XY)\) ):
$$
{\rm Ric}_i = \frac{5}{12}\ (i=1,2,3),\qquad {\rm Scal}=\frac{5}{2},\qquad {\rm Scal}^2=\frac{25}{4},\qquad |{\rm Ric}|^2=\frac{25}{24},\qquad |{\rm Riem}|^2=\frac{23}{12}.
$$
These follow from the general-chamber Ricci formula
$$
{\rm Ric}_1=\frac{x_1^2-x_2^2+6x_2x_3-x_3^2}{12\,x_1x_2x_3},\quad {\rm Ric}_2=\frac{-x_1^2+6x_1x_3+x_2^2-x_3^2}{12\,x_1x_2x_3},\quad {\rm Ric}_3=\frac{-x_1^2+6x_1x_2-x_2^2+x_3^2}{12\,x_1x_2x_3},
$$
specialized to \(x_1=x_2=x_3=1\) : each numerator becomes \(1-1+6-1=5\) (cyclically), each denominator \(12\) , giving \({\rm Ric}_i=5/12\) directly — a one-line hand check requiring nothing beyond substituting \(x_1=x_2=x_3=1\) . The scale-invariant ratios that survive the choice of normalization (identical whether computed in this Killing-form convention or in the physical \(R_6\) -normalization where \({\rm Ric}_i=1/(2R_6^2)\) , \({\rm Scal}=3/R_6^2\) ) are the load-bearing quantities:
$$
\frac{{\rm Scal}}{{\rm Ric}_i}=6=\dim K_6,\qquad \frac{|{\rm Ric}|^2}{{\rm Scal}^2}=\frac16,\qquad \frac{|{\rm Riem}|^2}{{\rm Scal}^2}=\frac{23}{75}=0.30\overline{6}.
$$
A reader reproducing \(\dim K_6 = 6\) from \({\rm Scal}/{\rm Ric}_i\) is performing a genuine consistency check, not a tautology: the ratio is computed from two independently-defined curvature contractions (a rank-2 and a rank-0 contraction of the same Riemann tensor), and its landing exactly on the integer \(\dim K_6\) — rather than some unrelated rational — is a certified identity of the specific homogeneous-space curvature tensor being used, not a normalization artifact. This ratio is exactly the sort of internal check that would fail immediately if a wrong curvature tensor (e.g., that of a different symmetric space) had been substituted.

 (b) The cubic invariants and the second Bianchi identity as a hard tripwire. Beyond the quadratic invariants, the weight-6 cubic curvature invariants at the same center are
$$
K_1 = R_{ab}{}^{cd}R_{cd}{}^{ef}R_{ef}{}^{ab} = -\frac{113}{72}, \qquad K_2 = R_{abcd}R_{aecf}R_{ebfd} = -\frac{5}{72}, \qquad |\nabla{\rm Riem}|^2=\frac14 \ne 0.
$$
The nonvanishing of \(|\nabla{\rm Riem}|^2\) is itself a checkable structural fact, not a residual error: it certifies that \(K_6\) is homogeneous but not locally symmetric ( \(\nabla{\rm Riem}\ne 0\) is precisely the failure of local symmetry), which is the geometric reason the graviton heat-kernel leg at order 6 carries an irreducible Gelfand–Tsetlin ladder term rather than vanishing identically as it would on a symmetric space. A reader can independently confirm zero second-Bianchi-identity violations by taking the cyclic sum \(\nabla_{[a}R_{bc]de}=0\) on the general-chamber Ricci/Riemann data and checking it holds identically at \(x_1=x_2=x_3=1\) ; this is the same check that flags the retracted value in §3 below.

 (c) Sphere heat-kernel calibration — an independent closed-form cross-check with no shared machinery. The heat-kernel convention used throughout, \(K(t)\sim(4\pi t)^{-d/2}\sum_k a_{2k}t^k\) (coefficient density per unit volume), obeys the exact product rule \(a_{2k}(M_1\times M_2)=\sum_{i+j=k}a_{2i}(M_1)a_{2j}(M_2)\) . On round unit spheres, where the Laplacian spectrum is known in closed form ( \(\ell(\ell+d-1)\) with textbook degeneracies), the corpus reports
$$
a_6(S^2)=\frac{4}{315},\qquad a_6(S^4)=\frac{74}{63},\qquad a_6(S^6)=\frac{1139}{63},\qquad a_6(S^6)_{\rm conformal}=\frac{5}{63},
$$
each obtained by two structurally independent routes — direct spectral summation of the exact eigenvalue/degeneracy data, and the general Gilkey universal heat-kernel formula evaluated on the sphere's known curvature invariants — agreeing to \(\sim4\times10^{-14}\) relative precision, i.e. to machine-arithmetic accuracy with no adjustable parameter linking the two routes. This calibration matters to UQF-3 specifically because it validates the same heat-kernel machinery (convention, product rule, Gilkey coefficient formula) that the still-open \(a_6^\partial\) boundary computation on \(K_6\) 's fixed locus will have to use; an engine that fails the sphere calibration could not be trusted on \(K_6\) either. A reader can redo this from scratch: sum \(\sum_\ell({\rm degeneracy}_\ell)\,e^{-t\,\ell(\ell+d-1)/R^2}\) for \(S^2\) ( \(d=2\) ) and \(S^4,S^6\) ( \(d=4,6\) ), expand the result in small \(t\) , and read off \(a_6\) as the coefficient of \(t^{3-d/2}\) — no number in this dossier needs to be assumed to perform that expansion.

 (d) The scale-free ratio \(a_4/a_2^2=66/125\) — a Levi-Civita-immune check. Because \(a_2\propto{\rm Scal}\) and \(a_4\) is a fixed linear combination of \({\rm Scal}^2\) , \(|{\rm Ric}|^2\) , \(|{\rm Riem}|^2\) (the standard Gilkey weight-4 combination, with the total-derivative pieces integrating to zero on the closed, boundary-free \(K_6\) ), the ratio \(a_4/a_2^2\) depends only on the dimensionless ratios already fixed in (a) above — \(|{\rm Ric}|^2/{\rm Scal}^2=1/6\) and \(|{\rm Riem}|^2/{\rm Scal}^2=23/75\) — with no dependence on \(R_6\) or any physical scale. A reader can verify \(66/125\) directly by substituting \({\rm Scal}=5/2\) , \(|{\rm Ric}|^2=25/24\) , \(|{\rm Riem}|^2=23/12\) into the standard Gilkey \(a_4\) scalar-bundle formula and dividing by \({\rm Scal}^2=25/4\) ; agreement with \(66/125\) is a cross-check that the same curvature data entering (a)–(b) above is self-consistent at the next heat-kernel order, independent of any physical normalization choice.

 (e) The Donnelly \(\mathbb{Z}_2\) equivariant defect — forced, not fitted. The orbifold boundary \(S^1_Y/\mathbb{Z}_2\) has exactly two isolated fixed points under \(\theta\mapsto-\theta\) ( \(\theta=0,\pi\) ), and the equivariant Donnelly trace is
$$
\sum_{\rm fixed\ points}\frac{1}{|1-dg|} = 2\times\frac{1}{|1-(-1)|}=2\times\frac12=1,
$$
an exact integer forced by the reflection acting as \(dg=-1\) on the normal direction at each isolated fixed point — there is no freedom to tune this; any \(\mathbb{Z}_2\) reflection with isolated fixed points gives this same combinatorics. Cross-checked against the product structure on \(S^2\times(S^1_Y/\mathbb{Z}_2)\) , the equivariant \(a_0\) -defect factorizes as \((1/2)\cdot a_6(S^2) = (1/2)\cdot(4/315)=2/315\) , verified by two independent computational routes agreeing to \(\sim10^{-5}\) (direct numerical peel) and to \(1.3\times10^{-14}\) (an independent Vandermonde-matrix route) — again with no shared free parameter between the two routes. A companion check, the twisted-circle trace \({\rm Tr}(\sigma e^{-tD})=1.000000000000\) exactly, evaluated at multiple \(t\) and found \(t\) -independent with a spectrum built from integer powers only, is the specific piece of evidence ruling out a competing wrong hypothesis: a genuine half-integer-power boundary heat-kernel tower (the signature of an ordinary hard Dirichlet/Neumann boundary rather than an equivariant \(\mathbb{Z}_2\) orbifold identification) would make this trace \(t\) -dependent. Its exact, flat constancy is what licenses treating \(S^1_Y/\mathbb{Z}_2\) as an orbifold reflection and redirects the open computation to the correct object — the mixed-boundary coefficient on \(K_6\) 's own fixed locus \(F\) — rather than a naive circle-boundary tower that does not in fact exist.

 (f) Grading-constant cross-check — pure linear algebra forced by \(D=13\) . The graded fibre operator entering the BRST/reflection grading is \(A={\rm diag}(1_{12},-1)\) on the 13-dimensional fibre — 12 even directions, 1 odd (reflected) direction, forced by the dimension count \(D=13\) together with the single reflected circle, not chosen freely. Its traces:
$$
{\rm tr}\,A = 12-1=11,\qquad {\rm tr}\,A^2 = 12\cdot1^2+1\cdot(-1)^2=13,\qquad {\rm tr}\,{\rm Sym}^2(A)=\frac{({\rm tr}\,A)^2+{\rm tr}\,A^2}{2}=\frac{121+13}{2}=67.
$$
This reproduces both headline weights exactly — the graded ghost weight \({\rm tr}\,\gamma_{\rm ghost}={\rm tr}\,A=11\) and the graded graviton weight \({\rm tr}\,\gamma_{\rm grav}={\rm tr}\,{\rm Sym}^2(A)=67\) — with the graded Block-A weight following as \(67-2(11)=45\) . A reader can redo this on one line: it needs only the eigenvalue multiplicities \((12,-1)\) fixed by \(D=13\) and the binomial-square identity for \({\rm tr}\,{\rm Sym}^2\) of a diagonal matrix; no external number is borrowed anywhere in the computation. The companion firewall check is that the ungraded \({\rm Sym}^2\) multiplet count on the same 13-dimensional space is a different number, \(91=\binom{14}{2}=(13^2+13)/2\) (set every sign to \(+1\) , i.e. \(A\to{\rm Id}_{13}\) , giving \({\rm tr}\,A=13\) ), and confusing \(67\) with \(91\) — or vice versa — is precisely the class of error this cross-check exists to catch. Both numbers are independently re-derivable by hand from the dimension count alone, and they must not agree with each other; a computation that outputs \(67\) where \(91\) belongs, or the reverse, has made a graded/ungraded bookkeeping error, not found a new number.

 Why these six cross-checks matter to a CERTIFIED-IRREDUCIBLE grade specifically. None of (a)–(f) closes R4 (the still-OPEN \(a_6^\partial\) coefficient) or R3 (the CERTIFIED-IRREDUCIBLE continuum wall) — that is not their purpose, and no dossier claims otherwise. Their purpose is to demonstrate that every piece of geometric machinery the eventual R4 computation will draw on is already independently verified, self-consistent, and reproducible by hand or by direct numerical summation, with the sphere calibration and the Bianchi-identity check functioning as the two hardest tripwires (an engine that gets either wrong cannot be trusted on the harder \(K_6\) boundary computation). This is what distinguishes "we have not yet computed \(a_6^\partial\) " (an honest, bounded, in-principle-finite gap) from "we do not know if our machinery works" (a much worse problem this section shows is not the case).

 3. Negative controls — frozen tripwires a correct re-derivation must NOT reproduce

 A negative control here is a specific value that a plausible-looking but wrong computation would produce, which the corpus has explicitly identified, retracted, and keeps live specifically so that any future re-derivation which lands on it is flagged as an error rather than mistaken for independent confirmation.

 \(|{\rm Riem}|^2(K_6) \ne 31/147\) . The correct, Bianchi-consistent value at the Killing-form center is \(23/12\) (scale-invariant ratio \(23/75\) ). The value \(31/147\) traces to a sign error in the Nomizu curvature-formula reduction and fails the second Bianchi identity check described in §2(b) above when tested directly — this is not a rival convention, it is an arithmetic inconsistency with a differential-geometric identity every correct homogeneous-space Riemann tensor must satisfy. Any re-derivation landing on \(31/147\) has reproduced the identified sign bug, not found an alternative correct answer. Numerically the retracted ratio is \(31/147 = 0.2108\ldots\) and the Bianchi-consistent replacement is \(23/75 = 0.3067\ldots\) — the same \(\sim31\%\) curvature-input correction logged separately in the " \(0.2109\) vs \(0.3067\) " negative control below (the corrected ratio raises the input, not lowers it), with the correctly-grounded companion ratio \(|{\rm Ric}|^2/{\rm Scal}^2 = 1/6\) unaffected by the fix. This tripwire is never revived. 

 \(|{\rm Riem}|^2(K_6) \ne 60\) . This is not a sign bug but a category error: \(60\) is the round-unit- \(S^6\) value, a different manifold with a different isometry group ( \(SO(7)\) transitive on \(S^6\) , versus \(SU(3)\) acting on the flag manifold \(K_6\) ). No metric rescaling or convention choice makes these two curvature invariants coincide; a computation landing on \(60\) has silently substituted the wrong homogeneous space's curvature data into a Gilkey-type universal formula.

 \({\rm tr}[a_6] \ne -2.817995812\times10^{94}\ {\rm GeV}^6\) . Retracted outright, not merely superseded: at the odd total dimension \(D=13\) , the relevant heat-kernel trace develops a half-integer pole with a power divergence that vanishes identically in dimensional regularization, meaning the object being assigned this numerical value is structurally ill-posed at this order in this scheme — it does not exist as a finite quantity to be quoted. Any dimensionful GeV \(^6\) figure attached to this trace should be read as an error signature, never as a stale-but-roughly-right number, and it is admissible only as a labeled scale-riding consistency coefficient, never as a physical positivity input.

 The claimed " \(-16/315\) two-route agreement" is REFUTED, not confirmed. A claimed cross-route match at this value was checked and found false; it must not be cited as supporting evidence for any downstream heat-kernel or positivity result.

 "Factorization reduces the problem" is RETRACTED, demoted to conjecture (= residual R6). The claim that the interacting base \(\times\) internal Euclidean measure factorizes — which would have made a constructive attack on the continuum limit tractable factor-by-factor — is not established; nonseparability is treated as load-bearing in this gate's own root-anchoring, and any re-derivation of R3/R6 that silently assumes factorization to simplify the continuum or boundary computation is smuggling back a retracted assumption, not making progress.

 Curvature input \(0.2109\) is wrong; the true value is \(0.3067\) ( \(=23/75\) ), a \(\sim31\%\) discrepancy. This is logged as a "half-right, half-wrong" internal-engine defect: the derivative-sector ratio in the same computation is confirmed correct, but the curvature input feeding the graviton a₆ engine is stale (pre-dating the \(31/147\to23/75\) fix in §2(b)). Any downstream magnitude computed with \(0.2109\) in place of \(23/75\) must be treated as untrustworthy until re-run with the corrected input — this is carried forward explicitly as an open action item (§ "Open holes," item 4), not silently patched.

 4. Exactly how a reader re-derives the result from scratch

 A reader with no access to anything beyond a differential-geometry and constructive-QFT background can reproduce every DERIVED and DERIVED/ESTABLISHED claim in this gate, and can independently verify the scope, location, and character of the one open wall, by working through the following steps in order.

 Step 1 — reconstruct the frozen branch's finite-cutoff kinematics. Fix the gauge content to \(SU(3)_c\times SU(2)_L\times U(1)_Y\) (always compact — no step in the frozen 13-dimensional construction ever introduces a non-compact gauge factor), place it on a Euclidean lattice with spacing \(a>0\) and finite volume \(L<\infty\) , and write the standard Wilson plaquette action in terms of compact-group link variables with Haar measure. No gauge-fixing is introduced at this stage — this is the single representational choice (§ derivation chain, step 1) that keeps the Gribov–Singer obstruction (Gribov 1978; Singer 1978; Neuberger 1987) from having any purchase on this leg, because that obstruction is specifically a statement about the failure of global gauge-fixing slices in the continuum, and there is no gauge-fixing here to fail.

 Step 2 — verify the Osterwalder–Seiler/Lüscher hypotheses by inspection. Check that the plaquette action is reflection-positive under time-slice reflection and that the link measure is the (positive) Haar measure on a compact group — both true by construction from Step 1. Osterwalder–Seiler (1978) then delivers the transfer operator \(T=e^{-aH_a}\) with \(0\le T\le1\) automatically; the equivalence \(T\le1 \Leftrightarrow H_a\ge0\) (for \(a>0\) ) is elementary spectral calculus and requires no separate argument — boundedness below is not a further assumption, it falls out of \(T\le 1\) directly. Apply the Perron–Frobenius theorem to the positive operator \(T\) to obtain a simple, non-degenerate vacuum eigenvalue. This reproduces Check 1 of §1 in full.

 Step 3 — reconstruct the strong-coupling gap. For \(\beta<\beta_{\rm conv}\) , set up the character/cluster expansion of Seiler (1982)/Münster (1981) and verify its convergence radius; within that radius the expansion supplies an explicit analytic lower bound on \(\Delta(a,L)\) . A reader wanting the honest scope boundary should also verify — by consulting the same lattice-gauge-theory literature — that \(\beta_{\rm conv}\) sits on the strong-coupling side of the deconfinement/continuum crossover, confirming that this gap is a genuine but different-regime result from the Clay continuum gap, exactly as disclosed in §1 Check 2.

 Step 4 — reconstruct the BRST quartet counting on the retained sector. List the retained field content \(\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\) , identify the BRST doublets (longitudinal/time-like gluon paired with ghost/antighost), and verify the alternating-sign quartet cancellation order by order in perturbation theory following Kugo–Ojima (1979). Note explicitly, as part of the reproduction, that this step's validity depends on \(Q^2=0\) , and treat that fact as an imported, audited claim rather than re-deriving BRST nilpotency here — the nilpotency proof itself, including the one open even-degree \(d=4\) boundary anomaly class, belongs to UQF-4, not to this reproduction.

 Step 5 — reconstruct the linearized graviton positivity count. Take the Lichnerowicz operator \(E_L\) on the transverse-traceless \(\mathrm{Sym}^2_0\) bundle over \(K_6\) at the Killing-form center, with certified spectrum \(\{1/6\ (\times6),\ 5/12\ (\times6),\ 7/6\ (\times6),\ 17/12\ (\times2)\}\) on the 20-dimensional TT bundle ( \({\rm tr}\,E_L=40/3\) , \({\rm tr}\,E_L^2=241/18\) , both directly summable from the listed eigenvalues and multiplicities: \(6\cdot\frac16+6\cdot\frac{5}{12}+6\cdot\frac76+2\cdot\frac{17}{12} = 1+\frac52+7+\frac{17}{6}=\frac{40}{3}\) , confirming the quoted trace by hand). Verify that exactly 2 of the 20 TT polarizations correspond to the physical transverse-traceless graviton degrees of freedom, matching van Nieuwenhuizen (1973) and Stelle (1978) — this reproduces the DERIVED-GIVEN-E (linearized) grade on this leg, with the explicit reminder that nothing here extends to the interacting or above-cutoff theory (that extension is the separately tracked residual R7).

 Step 6 — reconstruct the geometric cross-check lattice of §2 in full , following (a)–(f) above line by line: substitute \(x_1=x_2=x_3=1\) into the general-chamber Ricci/Scal formulas; compute the cubic invariants \(K_1,K_2,|\nabla{\rm Riem}|^2\) and verify the second Bianchi identity holds; independently sum the sphere heat-kernel traces for \(S^2,S^4,S^6\) from the closed-form Laplacian spectrum; verify \(a_4/a_2^2=66/125\) from the dimensionless curvature ratios; recompute the Donnelly \(\mathbb{Z}_2\) equivariant trace from the fixed-point data of \(\theta\mapsto-\theta\) ; and recompute the \(D=13\) graded-fibre traces \(11, 13, 67, 45, 91\) from the diagonal matrix \(A={\rm diag}(1_{12},-1)\) alone. Landing on every value quoted in §2, and clear of every value flagged in §3, is the complete reproducibility test this gate offers on its geometric side.

 Step 7 — locate, but do not attempt to close, the two remaining open objects. Confirm that the continuum wall (R3, the SL-5/ULFCB uniform large-field coercivity bound) requires a uniform bound \(\inf\langle\psi,H\psi\rangle\ge c'\Lambda_{\rm YM}>0\) through the \(a\to0\) , \(L\to\infty\) limit, that Bałaban's estimate \(\mu_n(L_n)\le e^{-c/g(2^na)^2}\) shows only that large-field configurations are measure-rare (not that they are dominated/excluded from hosting a soft physical mode), and that no numerical constants \(c',K,\kappa\) are supplied anywhere in this reproduction — supplying any such constant without a proof would itself be a fabrication. Separately, confirm that the \(a_6^\partial\) mixed Neumann \(\oplus\) Dirichlet boundary coefficient on \(K_6\) 's fixed locus \(F\) requires Levi-Civita transport (not the canonical-connection Casimir spectrum, which is exact through \(a_0,a_2\) but already off by \(1/24\) at the vector \(a_4\) order and therefore untrustworthy at \(a_6\) ) plus the Gelfand–Tsetlin ladder matrix elements connecting the 5 Weyl-inequivalent \(T^2\) -weight classes inside \({\rm Sym}^2_0\) — a finite, in-principle-computable but not-yet-enumerated calculation, sharing its open stratum with Gap-01. A reader completing Steps 1–6 has reproduced every banked result in this gate; completing Step 7 to an actual number would resolve R4 (the one UQF-3-specific open object) and would still leave R3 (the shared Clay-class continuum wall) exactly where it is — irreducible inside this framework, not inside this reader's competence.

 What "reproducing this gate" therefore means, precisely. Full reproduction is: (i) verifying three theorem-hypothesis checks by inspection (Steps 1–4), each landing at the tier the hypothesis-check actually supports; (ii) verifying one linearized-sector positivity count by direct trace summation (Step 5); (iii) independently regenerating an exact-rational geometric lattice of more than a dozen cross-checked quantities from the \(K_6=SU(3)/T^2\) root system and the \(S^1_Y/\mathbb{Z}_2\) fixed-point structure, matching every quoted value and none of the retracted ones (Step 6); and (iv) correctly locating, without attempting to fabricate a resolution for, the two named open objects — one shared with the field-wide Clay-class wall (R3), one specific to this gate's boundary geometry and finite in principle (R4) (Step 7). No hash, file, or unpublished intermediate result is needed anywhere in this procedure — every input is either a published theorem (Osterwalder–Seiler 1978; Lüscher 1977; Seiler 1982; Münster 1981; Kugo–Ojima 1979; Gribov 1978; Singer 1978; Neuberger 1987; van Nieuwenhuizen 1973; Stelle 1978; Bałaban c. 1984–89) or an exact rational computed from the \(A_2\) root system of \(SU(3)/T^2\) at the symmetric chamber center \(\vec u=(1,1,1)\) .

 Open gaps & the specialist closure path

 UQF-3's terminal is CERTIFIED-IRREDUCIBLE / RESOLVED +0 , read two-axis as RESOLVED-with-residual . That terminal is not a placeholder awaiting future work to overturn it; it is the honest name for a boundary this program has walked up to, characterized at full 13-dimensional precision on the frozen branch \(\mathfrak B_{\rm active} = [\mathcal M_4\times K_6\times S^2\times S^1_Y/\mathbb Z_2]_\times \oplus [\mathcal F^+_{\rm finite}\oplus\mathcal C_{\rm admiss}]_\oplus \otimes [\mathcal E_{\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\oplus\mathcal E_{\rm proton}]_\otimes\) , and cannot presently cross. What follows is the residual roster in full working detail — not a list of TODOs but a precise inventory of which analytic objects remain unconstructed, why each is hard in a way specific to it (never a generic "more work needed"), what a genuine close looks like stated target-blind, what a genuine refutation looks like (a negative result is also a legitimate terminal here), the machinery a specialist would actually pick up to start, and what else on the wider gate board moves if each one closes. Nine residuals are carried, each on its own row, none rolled into a hedge and none upgraded past its named disposition: R3 (the shared Clay-class wall itself), R4 (the finite \(a_6^\partial\) boundary heat-kernel object), the \(P(a_6)\ge0\) functional (the undecided sign-criterion sitting on top of R4), R1 (the \(d=4\) boundary anomaly, exported to UQF-4), R5 (nonperturbative confined-QCD IR positivity, exported to UQF-11), R6 (the full KK-tower inner product, demoted to conjecture), R7 (above-cutoff nonlinear graviton positivity, exported to UQF-14), R8 (interacting-branch OS \(\leftrightarrow\) BRST equivalence), and the GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM (whether local certificates glue to one global positive theory — theorem-debt, not an axiom).

 R3 — Constructive positivity of the full interacting 4D base (the shared Clay-class wall)

 (a) The precise open object. What does not yet exist, on this program's arena or on anyone else's, is the continuum limit of the finite-cutoff physical Hilbert space,
$$
\mathcal H_{\rm phys}^{\rm YM} \;\overset{?}{=}\; \lim_{a\to0,\,L\to\infty} \mathcal H_{\rm phys}^{(a,L)},
$$
constructed so that it is (i) a genuine, non-degenerate, positive-definite Hilbert space, (ii) supports a self-adjoint transfer-generator \(H=-\lim a^{-1}\ln T\) bounded below with a unique ground state, and (iii) does all this uniformly as the double limit \(a\to0,\,L\to\infty\) is taken — not merely at each fixed \((a,L)\) separately (that fixed-cutoff object is banked unconditionally; see the derivation-chain steps 1–3 in the main text, and is not in question here). In the explicit sub-lemma ladder this is the conjunction of SL-3 (does the limiting kernel exist as a positive-definite object at all — Gribov/Singer/Neuberger territory) and SL-4/SL-5 (Osterwalder–Schrader axioms H1–H3 plus the uniform large-field control bound H4, the "ULFCB": \(\inf\{\langle\psi,H\psi\rangle\}\ge c'\Lambda_{\rm YM}>0\) uniformly in \(n,a,L,\beta\) , equivalently a uniform link-measure bound \(\|\Phi_n\|_\kappa\le K\) ). This is not a restatement of the mass-gap problem: the SCOPE theorem banked in this program (derivation-chain step 7) proves positivity is a proper sub-wall of Gap-02, strictly narrower than "prove confinement" and running strictly one direction, gap-machinery \(\to\) positivity, never inverted. But SL-5 taken alone is identical in difficulty to the corresponding piece of the Clay problem, because the mechanism that would deliver it — a genuine, uniform, continuum kernel construction for marginal 4D Yang–Mills — is the same mechanism a mass-gap proof would also need. On the layered arena this sits entirely on the \(\times\) -Stage \(\mathcal M_4\) transfer-matrix direction; the compact fiber \(K_6\times S^2\times S^1_Y/\mathbb Z_2\) supplies the gauge group and matter content that the transfer matrix acts on but carries no time direction of its own, so R3 is a statement about the base factor only, not about the compact geometry.

 (b) Why it is hard, and the specific traps. The difficulty is not a missing clever trick; four-dimensional, asymptotically-free Yang–Mills sits at the one scaling class where the field's usual tools run out simultaneously. Three obstructions compound:

 The Gribov–Singer topological obstruction. Gauge-fixing (Lorenz/Landau or any standard choice) cannot be made global and smooth on the full non-abelian configuration space: Gribov (1978) exhibited gauge-equivalent configurations satisfying the same gauge condition, and Singer (1978) proved by an index-theoretic argument that no continuous global gauge-fixing section exists — the principal bundle of gauge orbits is topologically obstructed. BRST quantization, built by picking a slice and adding Faddeev–Popov ghosts to compensate, is unambiguous only in a perturbative neighborhood of the vacuum, where Gribov copies are parametrically distant. Once the path integral explores the full non-perturbative configuration space — which the continuum limit forces it to do — the Gribov region is unavoidable, and it is precisely the region where the naive BRST-cohomology positivity argument is not known to apply.

 The Neuberger \(0/0\) catastrophe. Neuberger (1987) sharpened this into a no-go for the most literal non-perturbative repair: any attempt to define a gauge-fixed lattice path integral via the naive Faddeev–Popov recipe produces exact cancellations between Gribov regions of opposite orientation, driving gauge-fixed lattice expectation values to \(0/0\) identically. This is why the field moved to gauge-invariant lattice constructions (Wilson-loop kinematics, no gauge-fixing at all) for the finite-cutoff result banked in step 2 of the derivation chain — but it also means there is no known non-perturbative BRST-based route to attack SL-3/SL-4 directly; whatever succeeds must look structurally different from "repair BRST at finite \(a\) and take a limit."

 Rarity is not domination. Bałaban's UV-stability program (CMP, roughly 1984–1989) proves the large-field region of the block-spin renormalization-group flow — the region whose fluctuations threaten the coercivity bound — is provably rare , with measure at RG step \(n\) obeying \(\mu_n(L_n)\le e^{-c/g(2^na)^2}\) , shrinking super-polynomially as blocks are refined. The specific trap is mistaking rarity for domination: ULFCB requires that even a rare region cannot host a soft, vacuum-orthogonal sequence carrying enough probability weight, in the delicate double limit, to violate the bound. A marginal, exactly-four-dimensional, asymptotically-free theory has no spare coercive margin — no small parameter to trade a slice of the rarity bound for the domination the bound needs, unlike super-renormalizable or IR-safe theories where such a parameter exists. Treating a rarity estimate as though it were already a domination proof is the single most common false step in this literature; a specialist should regard any purported closure that skips straight from a Bałaban-type measure bound to ULFCB without an explicit domination argument as unfinished, not complete.

 A second-order trap specific to writing this gate up honestly: do not invoke the free-field OS \(\leftrightarrow\) BRST equivalence (textbook — Weinberg QFT Vol. II, Schwartz Ch. 25, Henneaux–Teitelboim) as though it settles the interacting question. It is a theorem about a theory with no interaction vertices, and switching on interactions is exactly where the Gribov ambiguity and the large-field tails re-enter (tracked separately below as R8 ). Nor does numerical reproducibility of the exact-rational curvature and heat-kernel invariants used elsewhere in this arena (e.g. \(|{\rm Riem}|^2/{\rm Scal}^2=23/75\) agreeing to \(\sim10^{-14}\) across independent routes) constitute physical validation of positivity — reproducibility confirms arithmetic, not physics, and must never be presented as evidence toward R3.

 (c) What closes it, target-blind, and what a refutation looks like. A genuine close requires either (i) a new constructive-QFT idea that proves ULFCB directly — a uniform bound \(\|\Phi_n\|_\kappa\le K\) independent of \(n,a,L,\beta\) — most plausibly by finding some structural feature (an exact symmetry, an integrable substructure, a positivity-preserving RG flow) special enough to convert the existing rarity bound into a domination bound without needing a small coupling parameter; or (ii) a decidable negative result at finite, rigorously-controlled order — a proof, not a numerical hint, that a soft vacuum-orthogonal sequence survives some sub-limit of the double limit, refuting continuum positivity for this construction. That negative result would itself be an accepted, legitimate terminal — a refuting close, not a failure of the program to report. The success criterion for (i) is explicit and target-blind: a proof (not a numerical estimate, not a lattice simulation, however suggestive) that the bound holds uniformly through the \(a\to0,L\to\infty\) double limit, together with a demonstration that the resulting object satisfies the Osterwalder–Schrader axioms H1–H4 and reconstructs a genuine Wightman theory via Osterwalder–Schrader reconstruction. No numerical value for \(c',K,\kappa\) is asserted here, nor should any future closure attempt assert one without a derivation — supplying an unearned number would be exactly the fabrication this dossier refuses to commit. Neither a positive close nor a rigorous negative carries a favored outcome or an implied timeline; both are dispositions this gate accepts as terminal.

 (d) Machinery to start from. The starting toolkit is the constructive-QFT literature already banked in the derivation chain, used as a launchpad rather than a citation list: Osterwalder–Schrader reconstruction (the H1–H4 axiom system defining exactly what must be proved); Osterwalder–Seiler lattice reflection positivity (Ann. Phys. 110 (1978) 440) and Lüscher's transfer-matrix construction (CMP 54 (1977) 283) — the finite-cutoff base to be taken to a limit; Seiler's convergent expansions (LNP 159, 1982) and Münster (1981) for the strong-coupling analytic control that must somehow be extended past the crossover region; Bałaban's block-spin renormalization-group UV-stability sequence (the closest existing rigorous non-perturbative machinery, to be extended past the UV region into a large-field/IR-controlling estimate); and the Gribov–Singer–Neuberger literature, respected as a hard constraint on which classes of construction can possibly work (gauge-invariant, never gauge-fixed, at the non-perturbative level). A specialist also wants the general theory of coercive quadratic forms and soft-sequence non-existence (the abstract functional-analytic content of SL-0, unconditionally banked already) as the shape the eventual bound must take.

 (e) Leverage. This is the single highest-leverage object in the entire UQF-3 residual family, because of the SCOPE theorem: gap-machinery closes positivity as a byproduct, so any resolution of Gap-02 (the Yang–Mills mass gap) automatically resolves R3 and R5 (nonperturbative confined-QCD IR positivity), and substantially informs R6 (the KK-tower inner product, currently blocked precisely because R3 is open) and R8 (interacting OS \(\leftrightarrow\) BRST equivalence). It is also, definitionally, the same object as the interacting-sector content of the Clay Millennium Yang–Mills existence-and-mass-gap problem — a genuine resolution here is a resolution (or rigorous refutation) of one of the seven Clay Millennium problems, with everything that entails outside this program's scope as well as within it. Nothing else on the current gate board is gated on R3 closing — the board's two counted structural frontiers are Gap-13 and UQF-4, not this one — which is precisely why UQF-3 is correctly graded CERTIFIED-IRREDUCIBLE rather than left blocking other gates: the program has isolated the wall rather than let it propagate.

 R4 — The order-6 mixed boundary heat-kernel coefficient \(a_6^\partial\) on the fixed locus of \(K_6=SU(3)/T^2\) 

 (a) The precise open object. This is a genuinely finite, non-Clay-equivalent computation — the one UQF-3-specific open object not identical in kind to a community-wide unsolved problem, and the one with a concrete numerical target. On the totally geodesic fixed locus \(F\) of the \(\mathbb Z_2\) reflection \(\theta\mapsto-\theta\) acting on the active boundary \(S^1_Y/\mathbb Z_2\) inside the full 13-dimensional arena, the mixed Neumann \(\oplus\) Dirichlet heat-kernel coefficient at order 6 in the short-time expansion
$$
K(t) \sim (4\pi t)^{-d/2}\sum_k a_{2k}\,t^k
$$
is unmade. The published boundary heat-kernel literature (Gilkey-type mixed-boundary-condition coefficient tables) stops at \(a_5\) ; \(a_6^\partial\) on a mixed-boundary manifold of this type — with the fixed-locus structure of the \(A_2\) flag manifold \(K_6=SU(3)/T^2\) underneath it — has never been tabulated. Producing it requires completing the Gelfand–Tsetlin off-diagonal wall : the Lichnerowicz first-order (hopping) term on \(\mathrm{Sym}^2_0\) (the transverse-traceless graviton bundle, dimension 20) mixes the five Weyl-inequivalent \(T^2\) -weight classes, and the required off-diagonal connection matrix elements are \(SU(3)\) Gelfand–Tsetlin ladder matrix elements between adjacent GT patterns — exact in closed form via the standard lowering-operator formula, but not yet enumerated for this specific representation content. This owed object then feeds a coefficient \(c_3^\gamma\) via the Donnelly equivariant-trace relation, using the certified fixed-point data: two isolated fixed points \(\theta=0,\pi\) on \(S^1_Y/\mathbb Z_2\) , reflection \(g\) -trace \(\Sigma\,1/|1-dg| = 2\times1/|1-(-1)| = 1\) , per-fixed-point \(a_0\) defect \(+1/4\) (parity \(+\) ) and \(-1/4\) (parity \(-\) ), \(\det(I-d\sigma|_N)=2\) at each fixed point (reflection acts as \(-1\) on the normal 2-plane), fixed locus \(F\) totally geodesic with zero angle deficit, and \(\mathrm{Vol}(S^1_Y/\mathbb Z_2)=\pi R_Y = 1/(2M_U)\) at the chamber center.

 (b) Why it is hard, and the traps. This is a computation-debt, not a conceptual gap — but the trap is treating it as easier than it is because the surrounding scalar and vector data are already banked to high precision. The certified backbone is real and should be used as a sanity check, never mistaken for the answer: the scalar heat-kernel coefficients on \(K_6\) itself are known exactly ( \(a_0/a_0=1\) , \(a_2/a_0=5/12\) , \(a_4/a_0=11/120\) ; \(a_6/a_0\) itself remains OWED on the Gilkey constants even at the scalar level), and calibration values on round spheres are exact rationals cross-checked to \(\sim4\times10^{-14}\) relative precision: \(a_6(S^2)=4/315\) , \(a_6(S^4)=74/63\) (with \(a_6(S^6)=1139/63\) round-unit and \(a_6(S^6)_{\rm conformal}=5/63\) under the distinct conformal normalization, correctly kept separate), and the scale-free ratio \(a_4/a_2^2=66/125\) (Levi-Civita-immune). None of this touches the graviton-sector \(\mathrm{Sym}^2_0\) hopping term, which is the actual owed piece.

 Two independent computation routes exist and neither has produced the value. Route A (Gilkey/Lichnerowicz directly on \(\mathrm{Sym}^2_0\) , dimension 20) needs the certified Lichnerowicz spectrum — eigenvalues \(1/6\) (multiplicity 6), \(5/12\) (multiplicity 6), \(7/6\) (multiplicity 6), \(17/12\) (multiplicity 2), with traces \(\mathrm{tr}\,E_L=40/3\) , \(\mathrm{tr}\,E_L^2=241/18\) — plus the curvature operator \(\Omega=\mathrm{Riem}\) , but is blocked exactly at the owed GT hopping term. Route B (ghost + vector reconstruction) has the certified vector data ( \(E=\mathrm{Ric}=\tfrac{5}{12}\mathrm{Id}\) , \(\mathrm{tr}\,a_2=0\) , \(\mathrm{tr}\,a_4=-47/360\) ) and the scalar backbone \(a_6/a_2^3=7936/39375\) (banked across three or more independent engines), but its graviton leg is the same owed quantity. The two routes have not yet been made to agree on this leg, meaning there is presently no independent cross-check on whatever value eventually comes out. A specialist closing this must either find a third independent route or accept a single-route result with the honest caveat that it is uncross-checked until Route A and B converge. There is a documented cautionary precedent for exactly this kind of error: an internal \(a_6\) -engine curvature input of \(0.2109\) was found to be \(\sim31\%\) low against the certified true value \(23/75=0.3067\) — a half-right/half-wrong state (the derivative-sector ratio was confirmed correct, the curvature contraction input was not) that must be fixed and the computation re-run before any resulting magnitude is trusted; this is carried explicitly as a DISCLOSED-CORRECTED debt, not swept aside.

 The frozen negative controls that any candidate \(a_6^\partial\) value must respect, and that a specialist must never silently violate, are: \(|\mathrm{Riem}|^2(K_6)=23/12\) with ratio \(|\mathrm{Riem}|^2/\mathrm{Scal}^2=23/75\) — never \(31/147\) (a retracted Bianchi-violating Nomizu-sign-bug artifact from an earlier, wrong reduction) and never \(60\) (the value for the topologically distinct round \(S^6\) ); \(\mathrm{Scal}/\mathrm{Ric}_i=6=\dim K_6\) in both the \(R_6\) and Killing-form normalizations; \(|\mathrm{Ric}|^2/\mathrm{Scal}^2=1/6\) ; the second-Bianchi check \(|\nabla\mathrm{Riem}|^2=1/4\ne0\) certifying \(K_6\) is homogeneous but not locally symmetric (the structural reason the \(a_6\) graviton leg carries a Gelfand–Tsetlin ladder term at all — a locally symmetric space would not); and the cubic invariants \(K_1=R_{ab}{}^{cd}R_{cd}{}^{ef}R_{ef}{}^{ab}=-113/72\) , \(K_2=R_{abcd}R_{aecf}R_{ebfd}=-5/72\) . A candidate \(a_6^\partial\) computation that silently reintroduces \(31/147\) , or \(-2.817995812\times10^{94}\,{\rm GeV}^6\) (a separately RETRACTED quantity, structurally ill-posed at odd \(D=13\) — half-integer heat-kernel pole, power divergence, zero in dimensional regularization — never to be revived except as a labeled scale-riding consistency coefficient, never as a physical positivity input), anywhere in its intermediate contractions is producing a corrupted result, not a new answer, and should be rejected on sight regardless of the final number it yields.

 (c) What closes it, target-blind, and what a refutation looks like. Closing R4 means deriving the GT off-diagonal hopping matrix elements between the five Weyl-inequivalent weight classes explicitly, feeding them into the Lichnerowicz heat-kernel expansion on \(\mathrm{Sym}^2_0\) , and obtaining a definite exact-rational (or, at minimum, high-precision numerical) value for \(a_6^\partial\) , cross-checked between Route A and Route B to a precision comparable to the existing \(\sim10^{-14}\) agreement already achieved on the scalar/vector sector calibrations. The success criterion is explicit: a specific rational number for \(a_6^\partial\) , reproduced independently by both routes, that feeds a finite coefficient \(c_3^\gamma\) via \(\mathrm{tr}[a_6]^{\mathbb Z_2}=\tfrac12\,c_3^\gamma\) . This value is genuinely target-blind: no sign or magnitude is assumed in advance, and — critically — a definite-sign violation of the downstream \(P(a_6)\ge0\) functional (next residual) is an equally legitimate, accepted terminal outcome, a refuting close rather than a failure to find the "right" answer. What would count as evidence of a wrong or fabricated closure attempt: any claimed value not traceable step-by-step through the GT ladder-operator matrix elements and the certified Lichnerowicz \(E_L\) spectrum quoted above; any value that reintroduces \(31/147\) or the retracted \(\mathrm{tr}[a_6]\) number; or any claim that two-route agreement has been achieved without an explicit shown computation on both sides. Two prior claimed "agreements" in this exact neighborhood are named and withdrawn precisely so they are not mistaken for precedent: a purported " \(-16/315\) two-route agreement" is REFUTED/FALSE, not independent confirmation; and the \(K_6\) scalar color-factor value \(124/315\) for \(a_6/a_0\) , while dual-validated internally, is graded derived-PENDING-independent-target-blind-reproduction because the same engine that carries the R2/R4 computation-debt produced it — it must not be presented as a settled, clean win until reproduced by a genuinely independent route.

 (d) Machinery to start from. The concrete starting point is the standard \(SU(3)\) representation-theory toolkit: Gelfand–Tsetlin patterns and the explicit lowering/raising operator matrix elements for \(SU(3)\) irreps (standard GT ladder theory), applied to the specific \((p,q)\) content relevant to \(\mathrm{Sym}^2_0\) on \(K_6=SU(3)/T^2\) — the weight structure is already tabulated via the Peter–Weyl decomposition, with Casimir \(C_2(p,q)=(p^2+q^2+pq+3p+3q)/3\) and dimension \((p+1)(q+1)(p+q+2)/2\) (e.g. the adjoint \((1,1)\) : \(\dim=8\) , \(C_2=3\) , zero-weight multiplicity \(2\) ). The heat-kernel side is standard Gilkey mixed-boundary-condition machinery — the canonical coefficient-recursion reference class for manifolds with boundary and mixed Neumann/Dirichlet conditions — combined with the Donnelly equivariant heat-kernel trace formalism already in productive use for the \(\mathbb Z_2\) orbifold defect bookkeeping elsewhere in this arena (the \(K^+=\tfrac12 K_{\rm circle}+\tfrac12(\text{defect }+\tfrac12)\) , \(K^-=\tfrac12K_{\rm circle}-\tfrac12(\text{defect}-\tfrac12)\) orbifold-trace construction). A specialist should set up the Lichnerowicz operator \((E_Lh)_{ab}=\mathrm{Ric}_{ac}h^c{}_b+\mathrm{Ric}_{bc}h^c{}_a-2R_{acbd}h^{cd}\) explicitly in the GT basis, compute the off-diagonal blocks connecting adjacent weight classes, and run the standard \(a_6\) Gilkey coefficient recursion — a rational function of curvature contractions and the bundle data \(E,\Omega\) — through to completion, using the certified Einstein-center curvature invariants of Section 4 (this dossier's geometry basis) as the fixed inputs the recursion must be built from, never re-derived ad hoc.

 (e) Leverage. R4 feeds directly and only into the \(P(a_6)\) functional immediately below — it is the sole missing input to that decision criterion. It is explicitly shared with Gap-01 (named in the residual roster as a shared home), so a completed \(a_6^\partial\) value closes or substantially informs that gate simultaneously, at no extra cost beyond the computation itself. It does not touch R3/SL-5, and closing it does not move the Clay-class wall at all — this is the correct, sober scope: R4 is a real, finite, closable object, but it is not a backdoor into the interacting-continuum problem, and a specialist should resist any temptation to oversell an eventual R4 closure as progress on R3.

 The \(P(a_6)\ge0\) functional — the undecided sign criterion

 (a) The precise open object. Separate from R4 itself is the decision criterion built on top of it: a would-be local positivity functional \(P(a_6)\ge0\) , whose pass/fail semantics have not yet been formalized and whose sign cannot be evaluated until \(a_6^\partial\) (hence \(c_3^\gamma\) ) lands. Currently no sign is asserted, in either direction — stated explicitly to forestall the trap of assuming a pass in advance.

 (b) Why it is hard, and the trap. The trap here is not computational but logical-ordering . It would be easy to write \(P(a_6)\) down as though its non-negativity were the expected or default outcome (since the finite-cutoff sector is already known to be positive by the OS–Seiler/Lüscher theorem) and then treat any future \(a_6^\partial\) computation as merely filling in a formality. That ordering is backwards and is explicitly rejected: the one-sided pass semantics must be fixed before the sign is known, precisely so a violation cannot be quietly reinterpreted after the fact. A second trap is conflating this local, finite functional with the global continuum statement R3: passing \(P(a_6)\ge0\) would be a local consistency check, never a proof that the full interacting continuum theory is positive — the two are logically independent objects living at different scales of the construction (one lives entirely on the compact fiber's fixed-point structure; the other lives on the \(\mathcal M_4\) continuum limit).

 (c) What closes it, target-blind, and what a refutation looks like. Closing this means formalizing, in advance of evaluating the sign, the precise one-sided decision rule: a violation ( \(P(a_6)<0\) ) is a refutation of whatever local positivity structure the functional was designed to certify, while a pass ( \(P(a_6)\ge0\) ) establishes consistency only — it does not upgrade to a proof of anything beyond the finite local object it tests. The success criterion is an explicit, target-blind statement of this rule, written down and frozen before R4 supplies the numeric input, so the eventual evaluation is a mechanical sign-check against a pre-committed criterion rather than a post-hoc rationalization. A refutation — \(P(a_6)<0\) once \(a_6^\partial\) is known — is an entirely acceptable, informative terminal: it would say something concrete about where the local boundary construction fails, without impugning the separately-established finite-cutoff bulk result (Check 1 of the derivation chain).

 (d) Machinery to start from. This is largely a definitional/formal task rather than a new computation: specifying \(P(a_6)\) precisely in terms of \(c_3^\gamma\) and the surrounding admissibility chamber \(\mathcal F^+_{\rm finite}\oplus\mathcal C_{\rm admiss}\) conventions already fixed for this arena (selector v3, constraints C1–C14, the freeze-before-compare barrier), using the same Donnelly equivariant-trace bookkeeping ( \(\mathrm{tr}[a_6]^{\mathbb Z_2}=\tfrac12c_3^\gamma\) ) that governs the rest of the \(\mathbb Z_2\) -orbifold sector.

 (e) Leverage. Purely downstream of R4; it adds no new leverage of its own but is the mechanism by which R4's eventual numeric answer becomes a physics statement (a positivity certificate or a refutation) rather than sitting as an isolated number with no interpretive frame.

 R1 — The \(d=4\) boundary anomaly class (BRST/anomaly closure), inherited and exported to UQF-4

 (a) The precise open object. The retained-sector perturbative no-ghost certificate (derivation-chain step 4, Kugo–Ojima quartet cancellation) is conditional on the BRST charge being nilpotent, \(Q^2=0\) — an input audited from, and properly the responsibility of, UQF-4, not re-derived here; counting \(Q^2=0\) as a UQF-3 floor anchor would be a RELABEL_FAIL and is explicitly forbidden. Given \(Q^2=0\) , the specific open sub-piece is whether the associated anomaly class at boundary dimension \(d=4\) (on the \(\mathcal M_4\) factor, and separately relevant on \(K_6\times S^2\) ) vanishes. The relevant target-blind computation is a bounded bordism/ \(\eta\) -invariant calculation in the Dai–Freed framework. The bulk-vanishing result is already proved for \(d\le3\) (FOS Corollary 7.5); the \(d=4\) case is the open one, and the companion result (FOS Corollary 7.6) flags it as likely nonzero — a real, disclosed expectation, not a neutral "don't know."

 (b) Why it is hard, and the trap. The hard part is that this is a genuine index/anomaly computation, not a bookkeeping exercise: it requires the actual Dai–Freed eta-invariant or bordism-group computation for the specific bundle and boundary data of this arena at \(d=4\) , and the published boundary heat-kernel/anomaly tower (the same literature that stops at \(a_5\) for R4) does not currently extend a ready-made \(d=4\) answer. The trap, named explicitly, is to assume the class vanishes by analogy with the proved \(d\le3\) case, or because "the theory is anomaly-free by construction elsewhere" — that would be target-loading (assuming the convenient answer because the alternative is inconvenient) and is rejected outright. The companion result already disclosing "likely nonzero" makes the assumption-of-vanishing trap especially tempting and especially illegitimate here: a specialist must compute, not infer from a preference for a clean closure.

 (c) What closes it, target-blind, and what a refutation looks like. Closing R1 means carrying out the bounded target-blind bordism/eta computation for the single even-degree \(d=4\) boundary class and reporting whichever sign or magnitude it actually returns. A DERIVED-GIVEN-E close (per the row's own grading convention) means finding that the relevant class cancels — a real, checkable result, not an assumption. A refutation — finding the class is indeed nonzero, consistent with the FOS Cor. 7.6 expectation — is equally a legitimate, informative terminal: it would mean the retained-sector no-ghost certificate (R2, and by extension step 4 of the derivation chain) carries a genuine anomaly obstruction at \(d=4\) that must be independently cancelled by some other mechanism (e.g. a compensating boundary term or a global-structure fix) rather than being automatically consistent. Either outcome is reportable; assuming vanishing without the computation is the one move that is not permitted.

 (d) Machinery to start from. Standard Dai–Freed eta-invariant and bordism-group technology (the same machinery underlying the global center/anomaly/Pin cohomology chain used elsewhere in this arena — Smith normal form on the charge-character lattice, \(H^*(BPSU(3);\mathbb F_3)\) generators, the Milnor \(Q_1=\beta P^1-P^1\beta\) differential at \(p=3\) ), specialized to the \(d=4\) boundary stratum on \(\mathcal M_4\) and cross-checked against the certified \(d\le3\) FOS result as a consistency floor.

 (e) Leverage. This result is a genuine input to UQF-4's own certificate (it is exported there, not resolved here), and it feeds back into UQF-3 only insofar as UQF-4's \(Q^2=0\) result is consumed as an audited pointer by R2 (derivation-chain step 4). It does not touch R3, R4, or the continuum spine.

 R5 — Nonperturbative confined-QCD IR positivity, exported to UQF-11

 (a) The precise open object. Even granting a hypothetical resolution of R3 (continuum positivity for the base Yang–Mills sector), a further, IR-specific certificate is needed for the confined , strongly-coupled regime of QCD as it actually appears in the retained matter sector: a nonperturbative positivity statement in the deep infrared, where confinement dynamics (not just the UV-safe short-distance sector) determines the physical spectrum. This is explicitly named as needing IR data that cannot come from UV information alone — a UV-complete, positive short-distance theory does not by itself certify that the strongly-coupled IR reconstruction is also a genuine positive Hilbert space with the correct (confined, gapped) spectrum.

 (b) Why it is hard, and the trap. The trap is assuming that R3, if closed, automatically discharges R5 — it does not. R3 concerns whether the continuum Hilbert space exists and is positive at all; R5 concerns whether the specific IR sector this arena's matter content is confined into (real QCD-like confinement, including the mass-gap phenomenon in Gap-02's domain) has its own certificate. These are logically related (both live inside the SCOPE theorem's "gap-machinery" side) but are not the same statement, and a specialist should not treat an R3 closure as silently including R5.

 (c) What closes it, target-blind, and what a refutation looks like. Closure requires a nonperturbative IR construction — most plausibly riding on whatever machinery eventually resolves R3 and/or Gap-02, since the SCOPE theorem establishes that gap-machinery closes positivity as a byproduct — producing a genuine positive-definite confined-sector Hilbert space with the correct spectrum. A refutation would be a rigorous demonstration that the IR reconstruction fails to be positive-definite for this specific matter content, which would be a serious structural finding about the retained sector, not merely a technical gap.

 (d) Machinery to start from. This is explicitly exported to UQF-11 and is not attacked independently here; it inherits whatever machinery Gap-02/UQF-11 brings to bear on confined nonperturbative QCD (lattice-rigorous or constructive-QFT IR methods), plus anything R3 supplies once available.

 (e) Leverage. Automatically resolved as a byproduct if R3 (equivalently Gap-02) closes, per the SCOPE theorem. Until then it remains an open, separately-tracked export.

 R6 — Full KK-tower, all-loop positive-definite inner product

 (a) The precise open object. Whether the full interacting theory across the entire Kaluza–Klein tower (base \(\times\) internal coupling, all loop orders) carries a single positive-definite inner product. An earlier working assumption in this program — that the interacting measure factorizes as base \(\times\) internal, allowing the base-sector positivity result to be extended to the full tower by a simple product argument — has been explicitly retracted: "factorization reduce" is a named, RETRACTED move, now demoted to an open conjecture rather than a banked step. Non-separability of the interacting measure is treated as load-bearing, not a technicality to be waved past.

 (b) Why it is hard, and the trap. The trap is exactly the retracted move: assuming that because the free KK tower factorizes cleanly (each mode is an independent harmonic oscillator with its own positive norm), the interacting tower must too. Interactions couple KK levels to each other and to the zero-mode sector precisely through the vertices that make the theory interesting in the first place, and there is no a priori reason the resulting inner product remains a simple tensor product once those couplings are switched on. This residual is additionally blocked by R3 (the base-sector continuum construction is not yet in hand to even ask the question about the full interacting measure) and by UQF-9/UQF-10 (the internal-sector operator content those gates are responsible for is itself not yet fully closed).

 (c) What closes it, target-blind, and what a refutation looks like. A close requires either a genuine non-perturbative construction of the full base \(\times\) internal interacting measure shown to be positive-definite at all loop orders (almost certainly requiring R3 in hand first, since the base sector is the harder half), or a demonstration that some weaker, non-factorized but still positive structure survives even without literal factorization. A refutation — a rigorous demonstration that the interacting KK-tower measure is not positive-definite at some finite loop order or KK level — would be a serious, informative structural finding about this arena specifically (not a generic Clay-class statement), since it would identify a genuine internal inconsistency rather than an unproven idealization.

 (d) Machinery to start from. Whatever machinery eventually resolves R3 for the base sector, combined with the internal-sector operator/spectral technology under development for UQF-9/UQF-10 (KK mass spectra \(m^2_{(p,q),{\rm vec}}=(C_2(p,q)+\Delta_{\rm vec})/R_6^2\) and \(m^2_{(p,q),{\rm Dirac}}=(C_2(p,q)+\|\rho\|^2+\Delta_{\rm spin^c})/R_6^2\) with \(\|\rho\|^2=2\) , already tabulated for the free spectrum); the open task is the interacting extension of this data into an inner-product statement, not the free-spectrum bookkeeping itself.

 (e) Leverage. Blocked by, rather than feeding, R3; a downstream consumer of both R3 and the UQF-9/UQF-10 internal-sector program rather than an independent lever on either.

 R7 — Above-cutoff / nonlinear graviton positivity, exported to UQF-14

 (a) The precise open object. The banked graviton positivity result (derivation-chain step 6) is explicitly linearized : on the transverse-traceless \(\mathrm{Sym}^2_0\) bundle (dimension 20), exactly 2 of the modes are positive-norm physical polarizations, a clean statement about the free/linear graviton fluctuation around the background. What remains open is positivity for the nonlinear, above-cutoff graviton sector — i.e., once genuine graviton self-interactions and the full non-linear structure of the metric fluctuation are included, and once energies above whatever UV cutoff bounds the linearized treatment are reached.

 (b) Why it is hard, and the trap. The trap is treating the clean linearized count (2 positive-norm TT polarizations, a standard general-relativity fact reproduced here on the specific \(K_6\) background) as though it extends automatically to the interacting, above-cutoff regime. It does not: nonlinear gravity in an extra-dimensional embedding is exactly where UV-completion issues for gravity generically arise, and this residual is explicitly named as blocked by UQF-9 (the internal-sector completion gate) — the above-cutoff graviton question cannot be answered without first knowing what the internal-sector UV completion of gravity in this arena actually is.

 (c) What closes it, target-blind, and what a refutation looks like. Closure requires a genuine UV completion of gravity in this 13-dimensional arena (or a demonstration that whatever completion emerges preserves positivity for the graviton sector above the cutoff), a task that is definitionally part of UQF-9's remit, not UQF-3's. A refutation — a demonstration that some feature of the nonlinear or above-cutoff sector forces a negative-norm mode — would be a serious finding about the arena's gravity sector specifically.

 (d) Machinery to start from. Whatever UV-completion machinery UQF-9 brings to bear on the internal-sector gravity content; the starting certified data on this side is the free Lichnerowicz spectrum already tabulated ( \(1/6\times6\) , \(5/12\times6\) , \(7/6\times6\) , \(17/12\times2\) on \(\mathrm{Sym}^2_0\) ) as the linear starting point any nonlinear extension must reduce to.

 (e) Leverage. Exported to, and blocked by, UQF-9; not independently actionable from within UQF-3.

 R8 — OS reflection positivity \(\leftrightarrow\) BRST no-ghost equivalence, interacting active branch

 (a) The precise open object. The free-field theorem that Euclidean reflection positivity (Osterwalder–Schrader) and BRST cohomological no-ghost positivity are equivalent characterizations of the same physical positivity (textbook: Weinberg Vol. II, Schwartz Ch. 25, Henneaux–Teitelboim) is scoped strictly to the free field. Whether this equivalence survives once interactions are switched on is open, and is disclosed here as a corrected, demoted status: OS reflection positivity is downgraded from "the intrinsic positivity condition" to "a chosen sufficient Euclidean certificate" — a deliberate, disclosed correction of an earlier overstatement, not the originally-assumed intrinsic wall.

 (b) Why it is hard, and the trap. The trap is exactly the one flagged under R3(b): citing the free-field textbook equivalence as though it settles the interacting question. It settles nothing about the interacting branch, because the equivalence proof itself depends on structure (Fock-space mode decomposition, absence of vertices) that interactions destroy. This residual is stated as folding into R3 and R6 rather than standing fully independently, because the interacting-branch machinery that would establish (or refute) the equivalence is the same machinery those residuals need.

 (c) What closes it, target-blind, and what a refutation looks like. Closure requires either an interacting-branch proof that OS positivity and BRST no-ghost positivity remain equivalent (most plausibly as a corollary once R3's continuum construction and R6's full inner-product are in hand), or a demonstration that they diverge in the interacting theory — in which case the two conditions would need to be tracked and separately certified rather than treated as one, and the choice of which one is the physically fundamental condition would itself become an open physics question. Either finding is a legitimate, reportable terminal.

 (d) Machinery to start from. The same OS reconstruction and BRST-cohomology technology already in use for R3 and derivation-chain step 4, applied specifically to the question of whether the two constructions' physical Hilbert spaces coincide once interactions are present, rather than to either construction in isolation.

 (e) Leverage. Folds into R3/R6; not independently closable ahead of those, and not a lever on anything outside this residual family.

 The GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM — do local certificates glue to one global positive theory?

 (a) The precise open object. Every positivity result banked in this dossier is a local certificate: finite-cutoff base positivity (R3's known side), linearized graviton positivity (R7's known side), perturbative retained-sector no-ghost (R2). What has never been proved — and is explicitly named as theorem-debt rather than smuggled in as an assumption — is a GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM : that these local certificates, once all separately established, actually compose into one global positive theory across sectors, interactions, the KK tower, boundaries, the IR, and above-cutoff gravity simultaneously. "Local certificates compose to a global positive theory" is listed explicitly among the non-claims this gate must never print as proven.

 (b) Why it is hard, and the trap. The trap is the single most tempting shortcut in the entire residual family: treating the conjunction of R1 through R8 (were they all independently closed) as automatically equivalent to global positivity, on the intuitive but logically unjustified assumption that "if every piece is positive, the whole must be." This "compositionality" assumption is named explicitly as the converged blind assumption a specialist is most likely to reach for, and it is exactly the kind of unproven composition theorem that a rigorous QFT construction cannot take for granted — cross-sector interference terms, boundary contributions, and the interacting KK-tower non-factorization already identified in R6 are all concrete mechanisms by which local positivity could fail to add up to global positivity even if every individual local certificate were independently airtight.

 (c) What closes it, target-blind, and what a refutation looks like. Closure requires an actual composition theorem: a proof that positivity on each local certificate's domain, plus whatever gluing/consistency data relates them (the residual list itself — R1's boundary class, R4's fixed-point data, R6's non-factorized measure, R8's OS/BRST bridge), is sufficient to construct one global positive-definite Hilbert space for the full theory. A refutation would be a demonstration — via an explicit obstruction, perhaps an anomaly-type class or an interference term that does not cancel — that local certificates can be simultaneously satisfied while the global object fails to be positive, which would be a fundamental structural finding, not a technical setback.

 (d) Machinery to start from. This is presently unstarted as a technical program in its own right (it is named, not attempted) — the natural entry point is the general mathematical theory of gluing local quantum field theories (e.g. factorization-algebra or algebraic-QFT gluing frameworks that make "local certificates compose" into a checkable statement rather than an assumption), applied to the specific residual list R1–R8 as the concrete gluing data that would need to be shown consistent.

 (e) Leverage. This is the master theorem the entire residual family ultimately answers to: even a hypothetical simultaneous closure of R1, R3, R4/ \(P(a_6)\) , R5, R6, R7, and R8 would not, by itself, constitute a full closure of UQF-3's underlying physical question without this composition theorem also being established. Conversely, it is not needed to sustain the CERTIFIED-IRREDUCIBLE / RESOLVED +0 terminal — that terminal already honestly reflects a bounded, finite-cutoff result plus a named, irreducible continuum wall, without requiring the eventual full-theory composition to be resolved. The three-axiom floor for the full theory (AXIOM-ANCHOR, AXIOM-QUOTIENT \(H_{\rm phys}={\rm ker}\,Q/{\rm im}\,Q\) , and AXIOM-CONSTRUCTIVE-RECONSTRUCTION) is explicitly distinct from, and larger than, the two-axiom floor (AXIOM-STABILITY-OF-MATTER, AXIOM-BORN-SIGN) that suffices for the terminal graded here — a specialist should not conflate "what this gate's terminal requires" with "what the full research program would eventually need."

 Summary table — the residual family, at a glance

 Residual 
 Status carried forward 
 Blocks / blocked by 
 Exported to 

 R3 (continuum base positivity) 
 REDUCED-TO-AXIOM / CERTIFIED-IRREDUCIBLE 
 Highest leverage: closes R5, informs R6/R8 
 shared with Gap-02 (Clay) 

 R4 ( \(a_6^\partial\) boundary coefficient) 
 OPEN/UNMADE, no sign asserted 
 Feeds \(P(a_6)\) only; does not touch R3 
 shared with Gap-01 

 \(P(a_6)\ge0\) functional 
 OPEN/UNMADE, downstream of R4 
 Purely downstream of R4 
 — 

 R1 ( \(d=4\) boundary anomaly) 
 AUDIT (inherited), likely nonzero 
 Feeds retained-sector R2 conditioning 
 UQF-4 

 R5 (confined IR positivity) 
 OPEN 
 Auto-closes if R3/Gap-02 closes 
 UQF-11 

 R6 (full KK-tower inner product) 
 OPEN, demoted to conjecture 
 Blocked by R3 + UQF-9/UQF-10 
 — 

 R7 (above-cutoff graviton) 
 DERIVED-GIVEN-E (linear) / OPEN above cutoff 
 Blocked by UQF-9 
 UQF-14 

 R8 (OS \(\leftrightarrow\) BRST, interacting) 
 DISCLOSED-CORRECTED / OPEN, scoped 
 Folds into R3/R6 
 — 

 Composition theorem 
 UNPROVEN, named theorem-debt 
 Master theorem over all of the above 
 — 

 None of these nine rows is fabricated toward a value, none is silently assumed to pass, and none is rolled into a single hedge word. Each is a named, bounded, testable object with a stated target-blind success criterion and a stated refutation criterion — exactly the discipline the CERTIFIED-IRREDUCIBLE / RESOLVED +0 terminal is built to honestly support.

 Honest ceiling, scope & the endpoint

 0. What this section is for

 Every earlier section of this dossier argued for the certificate: the finite-cutoff
transfer-matrix theorem, the Kugo–Ojima quartet mechanism on the retained sector, the linearized
graviton's two positive-norm transverse-traceless polarizations, the exact-rational geometric
cross-checks, and the scope theorem that isolates positivity as a proper sub-wall of the Yang–Mills
mass-gap problem. This section does the complementary work on purpose. It draws the bright lines
around what has not been shown, prices every anchor the certificate actually spends, and then
writes the one endpoint sentence this gate — and no other sentence — is entitled to write. The
fixed grade is CERTIFIED-IRREDUCIBLE / RESOLVED +0 , two-axis read RESOLVED-with-residual .
Nothing below moves that grade; a reached terminal is a floor, not a ceiling, and this section's
entire purpose is to show why the floor is exactly where it is — not lower, and not artificially
higher.

 The discipline here is the same discipline used throughout the derivation chain: every disposition
below is written so that it stays true whether or not the still-open continuum leg eventually
resolves positive, resolves negative, or dissolves onto a different axiom than the one presently
adopted. Nothing is closed here by aiming at the desired positive-norm outcome and working
backward — that is precisely the target-loading failure mode this program is built to refuse, and
the highest-risk vector for it on this specific gate (the qualitative given-E matter content
supplying what must come out positive-norm, never a positivity certificate itself) is named
explicitly in §2 below rather than left implicit.

 1. What is explicitly NOT claimed

 1.1 "UQF-3 is closed." It is not, and no downstream summary of this dossier may say so.
CERTIFIED-IRREDUCIBLE names a terminal — a point at which the attack sequence legitimately stops
because it has run into an external wall shared by the entire field of constructive quantum field
theory — not a point at which the underlying mathematical question has been answered. The gate
carries a genuinely open continuum leg (§3 below, identical in kind to the interacting-sector
content of the Yang–Mills Clay Millennium problem); "closed" would assert that this leg no longer
exists as a live mathematical question, and that assertion is false. Every occurrence of this gate
elsewhere in this corpus must read CERTIFIED-IRREDUCIBLE, never "closed," never "solved," never
"proved."

 1.2 "Reflection positivity is derived," stated as an unqualified blanket claim. This is the
single most dangerous compression available and it is forbidden outright. The load-bearing
qualifier cannot be dropped: finite-cutoff Osterwalder–Seiler/Lüscher reflection positivity — at
fixed lattice spacing a>0 , fixed volume L<∞ , gauge-invariant Wilson-loop kinematics — is a real,
banked, DERIVED theorem, checked against Osterwalder–Seiler (1978), Lüscher (1977), Seiler (1982),
and Münster (1981), with no borrowed number anywhere in the chain. The fully interacting,
continuum-uniform statement obtained by removing both regulators simultaneously
( a→0, L→∞ ) is a separate, open, Clay-class object. Merging these two clauses into one unqualified
sentence — "reflection positivity holds," full stop — converts a checkable finite-volume theorem
into an unproven claim about a Millennium Problem, and no slide, abstract, or downstream section of
this corpus is permitted to make that merge.

 1.3 "Continuum positivity is solved." It is not solved, and it is important to be precise about
 what has actually happened to the continuum question rather than let a euphemism stand in for an
answer. The disposition assigned to the continuum leg is REDUCED-TO-AXIOM under the granularity
screen P1 (§4 of the derivation chain, §4 of the geometry-anchoring table above): the finite-cutoff
regime a>0, L<∞ is declared physically relevant, by axiom, and the strict a→0 continuum limit
is treated as an idealization whose uniform positivity is not adjudicated inside this gate.
Dissolution onto an axiom is not a synonym for solution. It is a change of question — from "does
the a→0 limit exist as a positive-definite Hilbert space, uniformly?" to "is the physically
relevant regime the finite-cutoff one, by declared axiom P1?" — and changing the question is not
the same act as answering the original one on its own terms. Any report that reads "dissolved onto
P1" and concludes "UQF-3 proves continuum positivity" has mis-stated the result and must be
corrected on sight.

 The negative control attached to this exact point is deliberately kept alive in the derivation
chain, and it matters here as much as it does there: a granularity move is not a universal solvent
that can be pointed at any residual to make it vanish. The order-6 mixed-boundary heat-kernel
coefficient a_6^∂ (R4, §3.4 below) does not live at a→0 — it is a fixed, finite-geometry
object on the totally geodesic fixed locus of the S¹_Y/ℤ₂ reflection, evaluated at the frozen
Killing-form center. Declaring an axiom about the continuum limit says nothing about it, and it
stays open regardless of which stance is adopted toward P1. The fact that this negative control was
checked and survived — i.e., the temptation to over-dissolve was tested and rejected — is part of
why the residual ledger below is trustworthy rather than a list quietly emptied by an all-purpose
axiom.

 1.4 "UQF-3 implies the Yang–Mills mass gap," in either direction. The dependency this gate
establishes is strictly one-directional: gap-machinery → positivity . The finite-volume spectral
apparatus — the SL-0 variational identity (kernel coercivity ⇔ no soft sequence ⇔ spectral gap,
a pure spectral theorem with no open content) and the SL-1 conditional spectral-conversion
statement ( Spec(H) ∩ (0,Δ) = ∅ with Δ ≥ c′Λ_YM for some structure constant c′ > 0 , with no
numeric value for c′ supplied or claimed) — is used in service of the positivity argument. It
is never used in the reverse direction to certify anything about confinement or the mass gap itself.
UQF-3's one genuinely new, target-blind structural contribution — the SCOPE theorem — is precisely
the proof that positivity does not require the mass gap : the continuum wall carried by this gate
(SL-3, SL-4, SL-5) is a proper sub-wall of the full Gap-02 mass-gap problem, strictly weaker as a
mathematical demand. Reversing the arrow — treating this gate's certificate, partial as it is, as
evidence toward Gap-02 — would be an invalid inference not licensed by anything derived here, and no
downstream text may make that inversion.

 1.5 "Given-E derives the certificate" / "selection substitutes for derivation." The observed
Standard Model matter content (given-E: three chiral generations, the electroweak-doublet/singlet
assignment, the Higgs Wilson-line sector) enters this gate exactly once, and only in one role: it
supplies the target spectrum — the list of fields whose norm must be checked — never a positivity
 certificate . Nothing about the fact that given-E was selected (rather than derived from first
principles elsewhere in this program) does any work toward proving that the resulting states carry
positive norm; positivity is checked on whatever spectrum given-E hands over, by the independent
machinery of Kugo–Ojima quartet cancellation and Osterwalder–Seiler/Lüscher reflection positivity.
This is the highest-risk target-loading vector on this specific gate and it is watched explicitly
for exactly this reason: it would be easy, and wrong, to claim that because the matter content
"came out chiral and anomaly-consistent elsewhere in the arena," its positivity is therefore
guaranteed by construction. It is not — positivity is a separate, non-trivial fact about the
BRST cohomology and the Euclidean correlator kernel, proved (where it is proved) by machinery
external to the shape-selection step, and left open (where it is open) by that same external
machinery, not by anything about how the spectrum was chosen.

 1.6 Dissolved is not solved, restated as a general principle, not just for P1. The point in
§1.3 generalizes: nowhere in this gate does a dissolution move (reframing a question so a prior
formulation no longer applies) substitute for an answer to the reframed question. The Shape-level
selection of the frozen branch (§1.7 below) is SELECTED-not-forced, and this gate's certificate is
checked on that branch, never used to argue for the branch. A dissolution changes what is being
asked; it does not, by itself, supply the missing proof, calculation, or refutation that the new
question still requires.

 1.7 "This gate closes, or depends on, Shape uniqueness." UQF-3 does not inherit, use, or in any
way depend on the corpus-level Shape-uniqueness screen W3, which remains independently OPEN at the
level of the whole program. The branch 𝔅_active = [M₄ × K₆ × S² × S¹_Y/ℤ₂] on which this gate's
certificate is checked is GIVEN — frozen, SELECTED-not-forced — and the positivity question is
asked and answered (where it is answered) conditional on that branch, never as part of an
argument that the branch itself is uniquely singled out. A reader must not infer, from anything in
this gate, that the frozen branch has been shown to be the only admissible geometry; that is a
separate, open question this gate is silent on by design.

 1.8 "Bałaban's UV-stability results supply the continuum construction." Bałaban's block-spin
renormalization-group program (roughly CMP, 1984–1989) is real, rigorous, and banked machinery, but
its scope is UV stability only. It proves that the large-field region of the RG flow — the region
whose fluctuations threaten coercivity — has measure μ_n(L_n) ≤ e^{−c/g(2ⁿa)²} , shrinking
super-polynomially under block refinement. It does not supply, and is not claimed here to
supply, a continuum construction of the physical Hilbert space, a proof of reflection positivity
surviving the double limit, or any piece of the mass-gap problem. This is a standing
misattribution risk flagged explicitly: citing Bałaban's UV-stability sequence as though it settled
SL-3/SL-4/SL-5 would overstate a real, narrower result into a false claim of a broader one. Rarity
of the dangerous region is not domination of it (§3.2 below); Bałaban proves the former, and the
latter — the actual content of the open wall — is not established by his results or by anything
else cited in this dossier.

 1.9 "The free-field Osterwalder–Schrader ↔ BRST equivalence proves the interacting case." The
textbook equivalence between Euclidean reflection positivity and BRST no-ghost cohomology
(Weinberg, Quantum Theory of Fields Vol. II; Schwartz Ch. 25; Henneaux–Teitelboim) is a genuine
DERIVED theorem, but it is a theorem about a theory with no interaction vertices. Switching on
interactions is exactly where the Gribov–Singer topological obstruction and the large-field tails
controlled (only in rarity, not in domination) by Bałaban's estimates re-enter. Step 5 of the
derivation chain is scoped free-field-only for this reason, and R8 in the residual ledger exists
specifically to track the fact that the interacting-branch version of this equivalence is not
established. No sentence in this dossier extends the free-field equivalence to the interacting
theory by citation alone.

 1.10 "Local certificates compose into a global positive theory." Nothing in this gate assumes,
proves, or is entitled to assume that certificates established locally — on the retained
perturbative sector, on the finite-cutoff base, on the linearized graviton, sector by sector —
automatically glue into a single global statement that the fully interacting theory (all KK modes,
all loop orders, all sectors simultaneously) is positive. This gluing step is named explicitly as
the GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM , and it is unproven : it is carried in this
dossier as a named theorem-debt, not smuggled in as an unstated assumption and not adopted as a
fourth axiom. "Compositionality" — the tacit assumption that local positivity results simply add up
— is identified in the corpus as a converged blind spot precisely because it is the kind of
assumption that feels too obvious to state and is exactly where an unearned closure would otherwise
sneak in. R6 (the full KK-tower, all-loop inner product) was previously carried as a "factorization
reduces the problem" claim; that claim has been retracted and the residual demoted to an explicit
open conjecture, non-separable, blocked on R3, precisely because the composition step it relied on
was never actually proved.

 1.11 "The d=4 boundary anomaly vanishes." No such claim is made, and the corpus's own
structural argument points the other way. The d≤3 bulk vanishing result is proven (FOS Corollary
7.5). The d=4 case (FOS Corollary 7.6) is open, and the mod-3 Milnor-operation analysis on the
classifying space BPSU(3) (the differential Q₁ = βP¹ − P¹β at the prime 3) shows that the class
 u₂ = 2y₁ + 2y₂ survives both the relevant primary differentials, which is a structural reason to
expect the d=4 obstruction nonzero , not a reason to expect it to vanish. Assuming vanishing
because the lower-dimensional case vanishes, or reviving the retired " d₃ = Sq³_ℤ " heuristic (which
is 2-primary and trivially annihilates 3-torsion, proving nothing about the surviving 3-primary
obstruction), is a named target-loading trap and is rejected here explicitly. This item (R1) is
carried OPEN, with a disclosed directional expectation, not a computed sign.

 1.12 "Hashes, reproducibility, or arithmetic cross-checks validate the physics." The
exact-rational identities in §7 of the derivation chain (the a_6 heat-kernel calibration family,
the graded-weight traces, the ℤ₂ defect products) are internal consistency checks confirming that
the arithmetic engine computing these geometric quantities is self-consistent and matches
independent calibration spaces ( S² , S⁴ , S⁶ ). They are not, individually or collectively,
evidence for or against reflection positivity of the interacting theory. Numerical reproducibility
of a curvature invariant to thirteen or fourteen decimal places is a statement about the arithmetic
pipeline, not a statement about whether H_phys is positive-definite in the continuum limit.

 2. The anchors paid — full accounting, nothing hidden

 UQF-3's certificate is bought, and its residual is priced, against a short, fully-enumerated list of
inputs. Naming every one of them — including the ones that cost nothing — is part of the honesty
discipline of this section.

 2.1 The two atomic axioms (the genuine floor). Exactly two value-free structural axioms are
consumed, and they are two representations of a single unitarity/contraction fact, not two
independent invariants:

 AXIOM-STABILITY-OF-MATTER — the Hamiltonian is self-adjoint, bounded below, and possesses a
 genuine ground state. This is the structural content that Osterwalder–Schrader reconstruction,
 where it succeeds, is a machine for producing; where the machine's continuum leg is not yet built
 (§3 below), this axiom is what is being asked to survive the limit.

 AXIOM-BORN-SIGN — probabilities are real and non-negative. This is the physical content of
 "no ghosts": every physical state, after the gauge-redundancy quotient, carries non-negative
 norm.

 These two are the entire atomic axiom floor for the retained, finite-cutoff part of this gate.
No other value-free primitive is spent to obtain the finite-cutoff theorem of §3.1 (below) or the
perturbative retained-sector quartet mechanism.

 2.2 The Q²=0 pointer — consumed, audited, not owned here. BRST nilpotency of the retained
perturbative quartet mechanism is consumed as an inherited result, carrying an AUDIT badge: it is
 used by Step 4 of the derivation chain (the Kugo–Ojima quartet argument), but it is not proved
inside this gate. It is a pointer exported to, and properly the responsibility of, UQF-4. Counting
 Q²=0 as an independent floor anchor of UQF-3 itself would be a relabeling error — it is imported,
checked at the seam, and re-exported, not minted here.

 2.3 Given-E — consumed as a target specification only. As detailed in §1.5 above, the observed
Standard Model matter content is consumed, but only as the qualitative list of what must be
checked for positive norm. It carries no quantitative weight in the certificate itself and is
watched explicitly as the highest-risk target-loading vector on this gate.

 2.4 The granularity axiom P1 — the price of the continuum disposition. The single axiom that
prices the disposition of the open continuum leg (rather than the retained finite-cutoff theorem)
is P1: the finite-cutoff regime a>0, L<∞ is declared the physically relevant regime, and the
strict continuum limit is treated as an axiom-conditional idealization. This is the axiom onto
which SL-5/ULFCB/R3 is dissolved ( REDUCED-TO-AXIOM ), and it is the only place in this gate where
a genuinely open mathematical question is converted into a terminal by declaration rather than by
proof. It is named here, priced here, and not hidden inside a roll-up verdict anywhere else in this
dossier.

 2.5 What costs nothing: the four irreducible dimensionful anchors. UQF-3 is scale-blind . It
consumes zero of the program's four irreducible dimensionful anchors —
 {M_Pl, α_i(M_Z), y_t, |V_us|} . These four numbers, which price every dimensionful prediction
elsewhere in this program, do not enter this gate's derivation chain at any step; the internal
radii and volumes appear only as shared frozen geometric background, never as tuned inputs to the
positivity argument itself. Consequently there are no σ-pulls associated with this gate — no
measured band is fit, so there is nothing to compare against a PDG value. The only quantitative
content of this gate consists of exact-rational geometric identities (§7 of the derivation chain)
checked against frozen negative controls ( |Riem|²/Scal² = 23/75 , never 31/147 ; |Riem|² = 23/12 ,
never 60 for the round S⁶ ), not measured bands. This is stated plainly so that "zero anchors
consumed" is not mistaken for "nothing was checked" — the checks that exist are exact-rational
structural identities, a different and in some ways more stringent kind of check than a
sigma-band comparison, but they are not anchor pulls and must not be reported as such.

 2.6 The three-axiom floor for the full theory, as distinct from the retained-sector floor. 
It is important to keep two different floors separate. The two-axiom floor of §2.1 covers what has
actually been shown (finite-cutoff RP, retained-sector perturbative quartet). If one instead asks
what axiom floor the entire, fully closed theory (interacting, all sectors, all loop orders, all
KK modes, continuum limit) would need, the corpus's compositionality finding identifies three:
AXIOM-ANCHOR, AXIOM-QUOTIENT ( H_phys = ker Q / im Q ), and AXIOM-CONSTRUCTIVE-RECONSTRUCTION (the
still-unproven claim that a full positive reconstruction exists across interactions, KK towers,
boundaries, the IR, and above-cutoff gravity simultaneously). The route from the present two-axiom,
finite-cutoff floor to that three-axiom, full-theory floor is the GLOBAL-ADMISSIBILITY-COMPOSITION
THEOREM (§1.10) — named as theorem-debt, not adopted as a fourth axiom, and not paid for anywhere in
this dossier.

 3. The wall itself, priced precisely: why this and not something smaller

 3.1 What is actually banked, unconditionally, at finite cutoff. At a>0 , L<∞ , with
gauge-invariant Wilson-loop kinematics (no gauge-fixing, hence no Gribov horizon in this leg), the
Osterwalder–Seiler/Lüscher construction delivers, as a genuine theorem rather than a plausibility
argument: a transfer operator T = e^{−aH_a} on H_phys^(a,L) satisfying 0 ≤ T ≤ 1 , a
self-adjoint H_a ≥ 0 , a Perron–Frobenius-unique vacuum, and a strictly positive finite-volume
spectral gap Δ(a,L) > 0 , computable in the strong-coupling expansion for β < β_conv (with the
divergence as β → β_conv disclosed as a caveat, not smoothed over). This is real, checkable, and
owes nothing to any number borrowed from elsewhere in the program.

 3.2 What is not banked: uniformity through the double limit. The single remaining leg is
whether H_phys^(a,L) converges, as a→0 and L→∞ simultaneously, to a genuine, non-degenerate,
positive-definite continuum Hilbert space supporting a self-adjoint generator bounded below with a
unique ground state — and whether that convergence is uniform , not merely valid at each fixed
 (a,L) separately. This is exactly the conjunction the derivation chain labels SL-3 (does the
limiting kernel exist as a positive object at all — Gribov/Singer/Neuberger territory) and
SL-4/SL-5 (Osterwalder–Schrader axioms H1–H3 plus the uniform large-field coercivity bound H4/ULFCB,
 inf⟨ψ,Hψ⟩ ≥ c′Λ_YM > 0 uniformly in n, a, L, β ). Three compounding obstructions make this hard
in a way specific to four-dimensional, asymptotically-free, marginal Yang–Mills, not a generic
"more work needed": the Gribov–Singer topological obstruction (no continuous global gauge-fixing
section exists on the full non-abelian configuration space, so BRST-based positivity arguments are
unambiguous only in a perturbative neighborhood the continuum limit forces the path integral to
leave); the Neuberger 0/0 catastrophe (naive Faddeev–Popov gauge-fixing on the lattice produces
exact cancellations between oppositely-oriented Gribov copies, ruling out the most literal repair);
and the rarity-is-not-domination gap in Bałaban's own UV-stability estimates ( μ_n(L_n) ≤
e^{−c/g(2ⁿa)²} proves the dangerous large-field region is rare, but a marginal, exactly
four-dimensional theory has no spare coercive margin — no small parameter — to convert rarity into
the domination ULFCB actually requires). None of the three obstructions is asserted here to be
insurmountable in principle; each is named as the specific, technical reason the wall has not yet
been crossed by anyone in the field, on this program's arena or any other.

 3.3 Why this is graded CERTIFIED-IRREDUCIBLE and not merely OPEN. The distinction matters and
is not cosmetic. An ordinary OPEN gate is a question this program has not yet gotten around to or
has a plausible attack line for that has simply not been executed. CERTIFIED-IRREDUCIBLE is
reserved for a question this program has walked up to, characterized completely at full
13-dimensional precision (the exact geometric constants of §4 of the derivation chain — Ricci
eigenvalues, |Riem|² , the cubic curvature invariants, the graded-weight traces — pin the object
being asked about with no ambiguity), and found to terminate at a wall that is (a) proven, not
merely suspected, to be shared by the entire field of constructive quantum field theory (it is,
concretely, the interacting-sector content of the Yang–Mills Clay Millennium Problem), and
(b) shown to have no lever available inside this framework that a generic attack on the same
problem elsewhere would not also need. Bałaban's rarity result and the Gribov–Singer–Neuberger
no-go results are not artifacts of this program's specific geometric arena; they are facts about
four-dimensional non-abelian gauge theory as such. A different choice of internal manifold, a
different squashing chamber, a different orbifold structure — none of it would remove this wall,
because the wall lives entirely on the M₄ factor, in the continuum limit of the transfer-matrix
construction, and does not depend on any property of K₆ , S² , or S¹_Y/ℤ₂ at all. That
independence from the internal geometry is exactly what licenses calling it irreducible rather than
merely unattempted.

 3.4 The one genuinely new, target-blind move: the scope theorem. It is worth restating why this
gate is not simply "the Clay problem under another name," because that would be an overclaim in the
other direction. The scope theorem — proper new content, checked target-blind — establishes that
reflection positivity is a proper sub-wall of the Gap-02 mass-gap problem: the dependency runs
strictly gap-machinery → positivity and never the reverse, so a hypothetical resolution of Gap-02
would carry R3 (and R5, and substantially inform R6 and R8) along with it, but a resolution of R3
alone would not by itself resolve Gap-02. This narrows, precisely, what this gate is actually
claiming to be irreducible: not "confinement is unsolved" in general, but specifically "uniform
continuum positivity, on its own terms, is exactly as hard as the corresponding piece of the mass
gap problem, and no easier." That is a sharper and more defensible claim than the undifferentiated
one, and it is why the gate can isolate its own wall — R3 — without that wall propagating and
blocking every other gate on the board. Concretely: the current gate board carries exactly two
counted structural frontiers, Gap-13 and UQF-4; UQF-3 is not among them, precisely because this
scope theorem lets the program treat R3 as a named, external, isolated wall rather than an
open dependency chain running back through the rest of the arena.

 4. The residual family, restated at the level of a scope table (not re-derived — priced)

 Nine residuals are carried on this gate's ledger, each on its own row, never rolled into a single
hedge and never upgraded past its named disposition. This section does not re-derive them (that
work is done in full elsewhere in this dossier); it re-states, for the purpose of an honest
endpoint, exactly what each one costs and what it does not touch.

 R1 (the d=4 boundary anomaly, §1.11) — AUDIT, inherited, exported to UQF-4; does not touch
 R3.

 R2 (quartet/no-ghost verified perturbatively only) — DERIVED-GIVEN-E at perturbative order on
 the retained sector; open non-perturbatively.

 R3 (constructive positivity of the full interacting 4D base) — the wall itself, §3 above;
 REDUCED-TO-AXIOM / CERTIFIED-IRREDUCIBLE.

 R4 ( a_6^∂ mixed-boundary heat-kernel coefficient) — OPEN/UNMADE, finite, in-principle
 computable, explicitly not the Clay wall, shared with Gap-01; no sign asserted.

 The P(a_6) ≥ 0 functional — OPEN/UNMADE, downstream of R4 only; a violation would refute a
 local structure, a pass would establish local consistency only, never a global proof.

 R5 (nonperturbative confined-QCD IR positivity) — OPEN, exported to UQF-11; an IR certificate
 cannot be manufactured from UV data.

 R6 (full KK-tower, all-loop positive-definite inner product) — OPEN, demoted to conjecture;
 base×internal factorization does not hold for the interacting measure; blocked by R3.

 R7 (above-cutoff, nonlinear graviton positivity) — DERIVED-GIVEN-E at the linearized level
 only; open above cutoff, exported to UQF-14, blocked on a UV completion of gravity.

 R8 (interacting-branch OS↔BRST equivalence) — OPEN/scoped; folds into R3/R6; the free-field
 equivalence (§1.9) is demoted to "a chosen sufficient Euclidean certificate," not the intrinsic
 wall.

 The composition theorem (§1.10) — UNPROVEN, named as theorem-debt, not assumed.

 Of these ten lines, exactly one — R4, and the P(a_6) functional built on top of it — is a
genuinely finite object, distinct in kind from the Clay-class wall, with a concrete
in-principle-computable numerical target and a named blocking stratum (the Gelfand–Tsetlin
off-diagonal hopping term connecting the five Weyl-inequivalent T² -weight classes in the graviton
 Sym²₀ bundle). Every other line either is the Clay-class wall (R3), is downstream of it (R5,
R6, R8), is inherited from a sibling gate (R1, exported to UQF-4), or is blocked on a UV completion
this program has not built (R7, exported to UQF-14). This is the honest shape of the residual
family: one small, finite, closable computation-debt sitting alongside one large, shared,
field-wide wall and its direct dependents.

 5. The closing endpoint statement

 Nothing left to attack inside this gate's own scope that has not been named, priced, and routed.
The finite-cutoff theorem is banked unconditionally. The retained-sector perturbative quartet
mechanism is banked conditionally on an audited, correctly-exported pointer. The scope theorem
isolating positivity from the mass gap is banked as genuine new content. The one remaining
Clay-class leg is characterized down to the specific technical obstructions (Gribov–Singer,
Neuberger, rarity-versus-domination) that make it hard, priced against exactly two atomic axioms
plus one granularity axiom, and dispositioned as a terminal this program shares with the whole
field rather than a gap unique to this arena. The one finite, program-specific object left on the
table (R4/ a_6^∂ ) is named as a bounded computation-debt, not a conceptual hole, with its blocking
stratum identified precisely (the GT off-diagonal ladder matrix elements) rather than gestured at.

 Nothing left. Anchored on: Shape: the frozen 13D branch 𝔅_active = [M₄ × K₆ × S² × S¹_Y/ℤ₂] ,
 K₆ = SU(3)/T² at the Killing-form Einstein center u=(1,1,1) , GIVEN and SELECTED-not-forced,
consumed only to supply the retained-sector target spectrum (given-E) and the totally geodesic
fixed locus on which the ℤ₂ reflection acts — never used to argue for its own uniqueness (W3 stays
independently OPEN and is not inherited here). Granularity: AXIOM P1 — the finite-cutoff regime
 a>0, L<∞ is declared the physically relevant regime; the continuum a→0 limit is
REDUCED-TO-AXIOM, an axiom-conditional idealization, not a solved theorem. Scale: UQF-3 is
scale-blind — zero of the four irreducible anchors {M_Pl, α_i, y_t, |V_us|} are consumed, and
there are no σ-pulls; the only quantitative content is exact-rational geometric identities
( |Riem|²/Scal² = 23/75 , Scal/Ric_i = 6 , the graded-weight trace tr γ_grav = 67 ) checked
against frozen negative controls, not measured bands. Observables: the finite-volume spectral gap
 Δ(a,L) > 0 and the transfer-operator bound 0 ≤ T = e^{−aH_a} ≤ 1 with H_a ≥ 0 , established
as genuine theorems at every finite (a,L) ; the two positive-norm transverse-traceless graviton
polarizations at linearized order; the retained-sector quartet cancellation verified
perturbatively. Dissolution: the continuum-uniform positivity question is not answered but
re-asked, converting "does the a→0 limit exist and stay positive?" into "is the finite-cutoff
regime the physically relevant one, by declared axiom?" — a dissolution onto granularity, checked
against a live negative control (the finite, a→0 -independent a_6^∂ object) that confirms the
dissolution is not being used as an all-purpose solvent. 

 Where this leaves a specialist, stated plainly rather than dressed up: the smallest remaining
object this gate can point to that is genuinely its own — not the shared Clay wall, not another
gate's exported responsibility — is the order-6 mixed Neumann⊕Dirichlet boundary heat-kernel
coefficient a_6^∂ on the totally geodesic fixed locus of K₆ = SU(3)/T² , blocked specifically at
the Gelfand–Tsetlin off-diagonal hopping term mixing the five Weyl-inequivalent T² -weight classes
in the transverse-traceless graviton bundle Sym²₀ . That single finite computation, once produced
target-blind by either of the two named routes (Gilkey/Lichnerowicz directly, or ghost+vector
reconstruction) and cross-checked between them, is what feeds the still-unformalized P(a_6) ≥ 0 
decision functional — whose answer, whichever sign it takes, is itself an accepted terminal, not a
step that must come out positive to count as progress. Beyond that one named, bounded, in-principle
computable object, everything else remaining on this gate's ledger is the shared, field-wide
Clay-class wall and its direct dependents, correctly routed to the gates (UQF-4, UQF-11, UQF-14)
whose scope they actually belong to, and correctly left, on this gate, as a limit on all present
knowledge rather than a defect of this specific program.

 Closure ledger — UQF-3 — reflection positivity

 Status (fixed): CERTIFIED-IRREDUCIBLE · RESOLVED +0

 The technical closure LEDGER (separate document)

 Gate: UQF-3 — reflection positivity / physical Hilbert space (BRST no-ghost + Osterwalder–Schrader positivity).
 Fixed grade (given, not re-derived here): CERTIFIED-IRREDUCIBLE / RESOLVED +0. PROMOTIONS:0.

 This ledger is the auditor's record: every object pinned at all three layers, every anchor's role stated, the full numbered derivation chain with exact values, the credit-ladder grade of each leg, the negative controls, and the endpoint line. No claim here exceeds what is printed in the narrative dossier; this document exists so a referee can check each atom independently.

 0. Layer-0 wall identity

 Wall statement. After stripping gauge redundancy from the frozen branch, does every physical state carry non-negative norm, and does the Euclidean path integral satisfy reflection positivity — equivalently, does a self-adjoint Hamiltonian bounded below with a genuine ground state exist on the interacting, continuum 4D non-abelian sector? This is the Osterwalder–Schrader/Osterwalder–Seiler constructive-QFT wall for 4D Yang–Mills, the same analytic difficulty class as the Clay Millennium existence-and-mass-gap problem (a distinct question sharing the identical construction).

 Wall class. External wall — not a gap manufactured by this framework's choices. It is a property of interacting relativistic quantum field theory in \(d=4\) that the entire mathematical-physics community has been unable to establish for any physically realistic non-abelian gauge theory since Osterwalder–Schrader axiomatized the problem (1973–75) and Wightman/Streater posed the constructive program. The frozen 13D branch inherits this wall on its retained 4D zero-mode sector; it does not create a new instance of it.

 Two sub-conditions on the same branch (no new object introduced): 
(a) BRST no-ghost. \(\mathcal H_{\rm phys} = \ker Q/\mathrm{im}\, Q\) contains only non-negative-norm states.
(b) Reflection positivity (Osterwalder–Seiler/Lüscher). The Euclidean correlator kernel is positive semi-definite, so the path integral defines honest probabilities and Osterwalder–Schrader reconstruction yields \(H \ge E_0 > -\infty\) with a genuine ground state.

 1. Layer-1 endpoint anchor

 The endpoint this gate is checked against is not a number but a theorem structure : existence of a positive-definite physical Hilbert space with a bounded-below, self-adjoint Hamiltonian, for the full interacting, continuum-limit 4D gauge/matter/graviton sector living on the frozen branch. The endpoint is binary and structural (exists / does not exist), not a fitted magnitude — so there is no PDG band, no \(\sigma\) -pull, and no anchor value to reproduce. The endpoint anchor is a theorem , and the ledger below shows exactly how much of that theorem is banked at finite cutoff and exactly which piece is the certified-irreducible external wall.

 Scope theorem (the one genuinely new, target-blind move produced by this gate). UQF-3's positivity requirement is a proper sub-wall of the Gap-02 mass-gap problem: reflection positivity does not require a mass gap to hold. The dependency runs strictly one way,
$$
\text{gap-machinery} \;\longrightarrow\; \text{positivity},
$$
never the reverse. This is a structural theorem about the logical relation between two Clay-class questions , not a numeric result, and it is banked as DERIVED .

 2. Layer-2 root stack

 2.1 Tier A — Shape / Scale / Granularity, full precision, all three layers pinned

 × Stage (metric geometry). Full frozen arena
$$
\mathfrak B_{\rm active} = \big[\mathcal M_4 \times K_6 \times S^2 \times S^1_Y/\mathbb Z_2\big] \times \;\oplus\; \big[\mathcal F^+ {\rm finite}\oplus\mathcal C_{\rm admiss}\big] \oplus \;\otimes\; \big[\mathcal E {\rm matter}\oplus\mathcal E_{\rm gauge}\oplus\mathcal E_{\rm Higgs}\oplus\mathcal E_{\rm proton}\big]_\otimes,
$$
 \(K_6=SU(3)/T^2\) (the full \(A_2\) flag manifold), \(D=4+6+2+1=13\) . The transfer matrix of the positivity construction lives on \(\mathcal M_4\) at finite lattice spacing \(a\) and finite volume \(L\) ; the compact fiber \(K_6\times S^2\times S^1_Y/\mathbb Z_2\) supplies the gauge group and matter content but carries no transfer-matrix time direction — it is inert for the positivity question itself and enters only through the target spectrum (given-E) and through the graviton/ghost grading (§2.3 below).

 ⊕ Rulebook. Gauge-invariant Wilson-loop kinematics (this choice removes the Gribov horizon from the banked finite-cutoff leg — no gauge-fixing is needed to state Wilson-loop reflection positivity); BRST gauge-fixed nilpotent operator \(Q\) with \(Q^2=0\) (imported as an AUDIT pointer from UQF-4, not re-derived here); Euclidean \(\to\) Lorentzian Osterwalder–Schrader reconstruction map; \(\mathbb Z_2\) parity table on \(S^1_Y/\mathbb Z_2\) (no-mirror structure, §2.3).

 ⊗ Actors. Transfer operator \(T=e^{-aH_a}\) acting on \(\mathcal H_{\rm phys}^{(a,L)}\) ; BRST operator \(Q_{\rm BRST}\) mapping off-shell fields to cohomology; Kugo–Ojima quartet operator on the longitudinal/time-like gluon \(\oplus\) ghost–antighost sector; Lichnerowicz endomorphism \(E_L\) acting on the graviton \(\mathrm{Sym}^2\) bundle.

 Shape root. GIVEN — the frozen branch is SELECTED-not-forced (Shape-uniqueness screen W3 is OPEN at corpus level). UQF-3 does not inherit W3 as closed: its certificate is checked on the branch, never used to select the branch. Shape supplies the routing \(SU(3)_c \leftarrow K_6\) , \(SU(2)_L \leftarrow S^2\) , \(U(1)_Y \leftarrow S^1_Y/\mathbb Z_2\) , and the orbifold fixed points \(\theta=0,\pi\) that drive both the reflection map (§2.3) and the still-open \(a_6^\partial\) boundary object (§4, R4).

 Granularity root — the axiom that carries the wall. Named axiom P1 : the finite-cutoff regime \(a>0\) , \(L<\infty\) is declared the physically relevant regime; the continuum limit \(a\to0\) is REDUCED-TO-AXIOM , i.e. the theory at strictly \(a=0\) is treated as an axiom-conditional idealization, not something solved inside this gate. This is the single root that carries the certified-irreducible residual: granularity is the lever that converts "wall" into "terminal" (§5).

 Scale root. UQF-3 is scale-blind : it consumes zero of the four irreducible dimensionful anchors \(\{M_{\rm Pl}, \alpha_i(M_Z), y_t, |V_{us}|\}\) . The frozen radii ( \(R_0=R_6=R_2=1.591549430918954\times10^{-17}\ {\rm GeV}^{-1}\) , \(R_Y = R_0/2 = 7.957747154594768\times10^{-18}\ {\rm GeV}^{-1}\) ) appear only as the shared background metric on which the transfer matrix and heat-kernel objects are defined; no dimensionful magnitude produced by UQF-3 is compared to a measured band.

 2.2 Full-precision geometric constants consumed as frozen negative controls

 All in the Killing-form normal metric, chamber center \(\vec u=(1,1,1)\) (exact rationals; identical in both metric normalizations for the scale-invariant ratios):

 Quantity 
 Exact value 

 \(\dim K_6\) 
 \(6\) 

 \(\mathrm{Ric}_i\) 
 \(5/12\) 

 \(\mathrm{Scal}\) 
 \(5/2\) 

 \(\mathrm{Scal}^2\) 
 \(25/4\) 

 \(\|\mathrm{Ric}\|^2\) 
 \(25/24\) 

 \(\|\mathrm{Riem}\|^2\) 
 \(23/12\) 

 \(\|\mathrm{Riem}\|^2/\mathrm{Scal}^2\) 
 \(23/75 = 0.3066666\ldots\) 

 \(\|\mathrm{Ric}\|^2/\mathrm{Scal}^2\) 
 \(1/6\) 

 \(\mathrm{Scal}/\mathrm{Ric}_i\) 
 \(6=\dim K_6\) 

 \(K_1=R_{ab}{}^{cd}R_{cd}{}^{ef}R_{ef}{}^{ab}\) 
 \(-113/72\) 

 \(K_2=R_{abcd}R_{aecf}R_{ebfd}\) 
 \(-5/72\) 

 \(\|\nabla\mathrm{Riem}\|^2\) 
 \(1/4\ne0\) (⇒ \(K_6\) homogeneous, NOT locally symmetric) 

 \(\chi(K_6)\) 
 \(6\) 

 \(\chi(S^2)\) 
 \(2\) 

 \(\chi(S^1_Y/\mathbb Z_2)\) 
 \(1\) 

 Frozen negative control (binding): \(\|\mathrm{Riem}\|^2/\mathrm{Scal}^2 = 23/75\) is the certified value — never \(31/147\) (a retired Nomizu-sign-bug artifact) and \(\|\mathrm{Riem}\|^2\) is never \(60\) (that is the round-unit \(S^6\) value, a different space). Any \(a_6^\partial\) computation that lands on \(31/147\) or on \(60\) has reproduced the withdrawn artifact, not the geometry, and must be rejected before its output is read.

 2.3 The \(\mathbb Z_2\) / grading data feeding the ghost and graviton sectors

 Donnelly \(\mathbb Z_2\) equivariant defect on \(S^1_Y/\mathbb Z_2\) (reflection \(\theta\mapsto-\theta\) , two isolated fixed points \(\theta=0,\pi\) ):
- \(g\) -trace: \(\sum 1/|1-dg| = 2\times 1/|1-(-1)| = 2\times 1/2 = 1\) .
- Per-fixed-point \(a_0\) defect: \(+1/4\) (even/+ parity), \(-1/4\) (odd/ \(-\) parity).
- \(\det(I-d\sigma|_N)=2\) at each fixed point (reflection acts as \(-1\) on the normal 2-plane).
- Fixed locus \(F\) totally geodesic, zero angle deficit.
- \(\mathrm{Vol}(S^1_Y/\mathbb Z_2)=\pi R_Y = 1/(2M_U) = 5.000000000000000\times10^{-17}\ {\rm GeV}^{-1}\) at chamber center.
- Twisted-circle trace \(\mathrm{Tr}(\sigma e^{-tD}) = 1.000000000000\) exactly, \(t\) -independent — integer powers only, which rules out a half-integer boundary tower and certifies that the naive "boundary \(a_6\) " object some earlier passes reached for is the wrong object.

 Chirality index on \([0,\pi]\) : \(n_L=+3\) , \(n_R=0\) — three chiral families, no mirror. This is consumed as the source of the given-E target spectrum (§3, anchor table) — it tells the gate what content must come out positive-norm, never supplies a positivity certificate itself.

 Graded fibre weights ( \(A=\mathrm{diag}(1_{12},-1)\) on the 13-dimensional fibre, the reflection grading operator):
$$
\mathrm{tr}\,A = 11,\qquad \mathrm{tr}\,A^2 = 13,
$$
$$
\mathrm{tr}\,\mathrm{Sym}^2(A) = \frac{\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2}{2} = \frac{13+121}{2} = 67.
$$
Graded graviton weight \(\mathrm{tr}\,\gamma_{\rm grav} = 67\) ; graded ghost weight \(\mathrm{tr}\,\gamma_{\rm ghost} = 11 = \mathrm{tr}\,A\) ; Block-A graded weight \(= 67-2(11) = 45\) , forced by \(D=13\) plus the reflection grading.

 Firewall (binding, never violate): the graded graviton weight is 67 , not the ungraded \(\mathrm{Sym}^2\) multiplet count 91 . These are two distinct quantities computed from the same fibre by two different traces (graded vs. ungraded); a dossier or ledger that swaps them has committed exactly the kind of fabrication error this ledger exists to prevent.

 2.4 Tier B screens

 Screen 
 Verdict 
 Consequence for UQF-3 

 W3 (Shape uniqueness) 
 OPEN at corpus level 
 UQF-3 does not depend on it; the positivity certificate is checked on the given branch, never used to argue the branch is forced 

 P1 continuum-reframe (BS-1 class) 
 \(a\to0\) continuum-uniform positivity is a continuum unicorn \(\to\) REDUCED-TO-AXIOM 
 This is the granularity move that converts the open construction (R3) into a named axiom-conditional wall rather than a raw unsolved gap 

 Negative control on P1 
 PASSED : a granularity attack on a finite wall buys nothing 
 The finite \(a_6^\partial\) + \(\mathbb Z_2\) defect computation (R4) is explicitly NOT dissolved by invoking P1 — it stays on the open roster as a genuinely finite, in-principle-computable object 

 Anchor-vs-target gate 
 UQF-3 consumes only value-free measured anchors; may not be closed on any quantity it predicts 
 Highest-risk target-loading vector is given-E (the observed SM matter content), which supplies the target spectrum only — watched explicitly, never used as a positivity certificate 

 3. Measured / structural anchors — role table

 UQF-3 consumes none of the four dimensionful anchors. Its inputs are value-free structural axioms plus one inherited pointer.

 Anchor 
 Type 
 Role 
 Disposition 

 AXIOM-STABILITY-OF-MATTER 
 value-free structural axiom 
 Hamiltonian self-adjoint, bounded below, ground state exists 
 CONSUMED — atomic floor anchor 

 AXIOM-BORN-SIGN 
 value-free structural axiom 
 probabilities real and non-negative 
 CONSUMED — atomic floor anchor, independent of stability 

 given-E (observed SM matter content) 
 qualitative target, not a number 
 supplies what content must come out positive-norm; never a positivity certificate itself 
 CONSUMED as target spec only — highest-risk target-loading vector, watched 

 \(Q^2=0\) (BRST nilpotency) 
 inherited result 
 feeds the Kugo–Ojima quartet mechanism (derivation step 4) 
 CONSUMED, AUDIT badge — pointer exported to/from UQF-4; counting it as a UQF-3-owned floor anchor is a RELABEL_FAIL 

 \(\{M_{\rm Pl}, \alpha_i, y_t, \lvert V_{us}\rvert\}\) 
 four irreducible dimensionful anchors 
 — 
 NOT CONSUMED — UQF-3 is scale-blind 

 \(\sigma\) -pulls: none. No dimensionful magnitude is fitted by this gate, so there is no PDG-band pull to report. The only quantitative cross-checks are exact-rational geometric identities matched against the frozen negative controls of §2.2 and §4 — a consistency check, not a measurement comparison.

 Axiom floor (atomic, exactly two): AXIOM-STABILITY-OF-MATTER + AXIOM-BORN-SIGN. Bounded-below- \(H\) , positive-norm states, and Born-rule sign positivity are three faces of one contraction/unitarity fact, not three independent invariants — counting them as three would double/triple-count the floor.

 Wall object (open construction, explicitly not a floor anchor): AXIOM-RECONSTRUCTION-BRIDGE-OPEN — the statement that a full interacting Euclidean measure exists and reconstructs stability plus Born-sign in the continuum. This is co-extensive with the 4D-Yang–Mills Clay wall (residuals R3/R6 below), the \(S^1/\mathbb Z_2\) boundary object (R4), and the UQF-11 IR sector (R5). It is the open construction the gate is honest about, not an atom of the floor.

 4. Derivation chain — numbered ledger, each step with its exact value and grade

 # 
 Step content 
 Exact value / statement 
 Grade 

 1 
 Finite-cutoff setup: \(a>0\) , \(L<\infty\) , gauge-invariant Wilson-loop kinematics (no gauge-fixing \(\Rightarrow\) no Gribov horizon on this leg) 
 — (setup) 
 setup 

 2 
 Osterwalder–Seiler / Lüscher reflection positivity \(\Rightarrow\) positive-definite \(\mathcal H_{\rm phys}^{(a,L)}\) 
 \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) 
 DERIVED 

 3 
 Perron–Frobenius \(\Rightarrow\) simple (unique) vacuum + strict finite-volume gap; strong-coupling expansion computes it for \(\beta<\beta_{\rm conv}\) (divergence as \(\beta\to\beta_{\rm conv}\) disclosed) 
 \(\Delta(a,L)>0\) ; explicitly not the Clay mass gap — finite-lattice gap can collapse as \(a\to0\) ("impostor inflation" not committed) 
 DERIVED/ESTABLISHED 

 4 
 Retained-sector BRST/Kugo–Ojima quartet mechanism removes longitudinal + time-like gluon polarizations against the ghost–antighost pair 
 \(\mathcal H_{\rm phys}=\ker Q/\mathrm{im}\,Q\) 
 CERTIFICATE-CONDITIONAL on UQF-4's \(Q^2=0\) 

 5 
 Free-field Osterwalder–Schrader \(\leftrightarrow\) BRST equivalence (textbook result) 
 — 
 DERIVED , scoped free-field only 

 6 
 Linearized graviton positivity: transverse-traceless \(\mathrm{Sym}^2_0\) carries exactly its physical polarizations 
 \(\dim\,\mathrm{Sym}^2_0=20\) ; TT sector has 2 positive-norm polarizations 
 DERIVED-GIVEN-E (linearized) 

 7 
 Scope theorem: positivity proven a proper sub-wall of the mass-gap problem 
 \(\text{gap-machinery}\to\text{positivity}\) , never inverted 
 DERIVED (the new target-blind move) 

 8 
 Continuum spine SL-0 … SL-5 (see §5) 
 — 
 mixed BANKED / OPEN 

 9 
 Bałaban UV-stability results — scoped narrowly 
 UV stability only ; does not supply continuum construction, RP, or the mass gap (standing misattribution risk explicitly flagged) 
 BANKED but scoped 

 10 
 \(a_6^\partial\) mixed Neumann \(\oplus\) Dirichlet boundary heat-kernel coefficient on the totally-geodesic fixed locus of \(K_6=SU(3)/T^2\) 
 finite object, UQF-3-specific, NOT the Clay wall 
 OPEN/UNMADE , no sign asserted 

 11 
 \(P(a_6)\ge0\) decision functional (built from step 10) 
 violation \(\Rightarrow\) refute; pass \(\Rightarrow\) consistency only, never a positive proof 
 OPEN/UNMADE , downstream of step 10 

 Reading the ladder. Steps 1–3 and 7 are fully banked theorems against published literature (Osterwalder–Seiler, Ann. Phys. 110 (1978) 440; Lüscher, CMP 54 (1977) 283; Seiler, LNP 159 (1982); Münster (1981)) with no borrowed numbers — every quantity in steps 1–3 ( \(0\le T\le1\) , \(H_a\ge0\) , \(\Delta(a,L)>0\) ) is a structural inequality, not a fitted magnitude. Step 4 is conditional on an external gate (UQF-4) and is graded accordingly — it is not re-derived or assumed here. Steps 10–11 are the two live open items that are genuinely specific to this gate (not shared with the Clay wall) and are carried honestly as OPEN, not folded into the certified-irreducible verdict.

 5. The continuum spine SL-0 … SL-5 — exactly where the wall sits

 Stage 
 Content 
 Status 

 SL-0 
 Variational identity: kernel coercivity \(\Leftrightarrow\) no-soft-sequence \(\Leftrightarrow\) spectral gap (pure spectral theorem) 
 BANKED , no open content 

 SL-1 
 \(\mathcal M_4\) spectral conversion: if uniform exponential clustering holds, then \(\mathrm{Spec}(H)\cap(0,\Delta)=\emptyset\) with \(\Delta\ge c'\Lambda_{\rm YM}\) for some structure constant \(c'>0\) and dynamically generated scale \(\Lambda_{\rm YM}\) 
 BANKED — derives the logical structure only; \(c'\) is explicitly not supplied (supplying a number here would be fabrication) 

 SL-2 
 Finite- \(a\) base = derivation steps 1–3 
 BANKED/ESTABLISHED 

 SL-3 
 Continuum kernel construction: does \(\mathcal H_{\rm phys}^{\rm YM}\) remain a genuine positive Hilbert space as \(a\to0\) ? 
 OPEN — the prior question. Gribov/Singer/Neuberger: no gauge-fixing-based positive-kernel construction is known to survive the Gribov horizon in the continuum 

 SL-4 
 Wightman/Osterwalder–Schrader axioms H1/H2/H3 — measure existence, RP survival, nontriviality 
 OPEN — Clay-class 

 SL-5 
 H4 = uniform large-field coercivity bound ("ULFCB"): \(\inf\{\langle\psi,H\psi\rangle\}\ge c'\Lambda_{\rm YM}>0\) uniformly through the continuum limit, equivalently a link-measure bound \(\|\Phi_n\|_\kappa\le K\) uniform in \(n,a,L,\beta\) 
 OPEN — THE CLAY CORE, THE WALL — dispositioned REDUCED-TO-AXIOM via granularity screen P1. No constants \(c',K,\kappa\) supplied 

 The precise obstruction at SL-5. Bałaban's analysis shows the large-field region of the RG flow is rare : at RG step \(n\) , \(\mu_n(L_n)\le e^{-c/g(2^na)^2}\) (Gaussian-type suppression, structure constant \(c\) not fitted to any number here). But rarity is not domination — ULFCB requires that the rare large-field directions be unable to host a soft, vacuum-orthogonal physical sequence, and marginal \(d=4\) Yang–Mills theory has no spare coercive margin to supply that for free. This is the exact mathematical shape of the wall: a known-rare set of field configurations that has never been shown to be powerless , in a theory that is exactly on the marginal boundary where a small failure of domination is enough to break the bound.

 6. Credit-ladder grading of every leg

 Leg 
 Credit-ladder grade 
 Basis 

 Finite-cutoff transfer matrix ( \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) ) 
 DERIVED 
 Published theorem (Osterwalder–Seiler/Lüscher), applied without modification to Wilson-loop kinematics on the frozen branch 

 Finite-volume unique vacuum + strict gap \(\Delta(a,L)>0\) 
 DERIVED/ESTABLISHED 
 Perron–Frobenius + strong-coupling expansion, disclosed convergence caveat 

 Quartet/no-ghost retained sector 
 DERIVED-GIVEN-E (perturbative) 
 Conditional on UQF-4's \(Q^2=0\) (imported, AUDIT badge); given-E supplies target content 

 Free-field OS \(\leftrightarrow\) BRST equivalence 
 DERIVED , scoped free-field 
 Textbook result; explicitly not extended to interacting branch 

 Linearized graviton TT positivity 
 DERIVED-GIVEN-E (linearized) 
 Consumes given-E for matter content; graviton sector itself is a pure linearized-gravity computation 

 Scope theorem (positivity \(\subsetneq\) mass gap) 
 DERIVED 
 New target-blind structural result, no numeric content, no anchor consumed 

 SL-0, SL-1, SL-2 
 BANKED 
 Spectral theorem / conditional structural bound / finite-cutoff base 

 SL-3, SL-4 
 OPEN 
 Genuinely unresolved community-wide questions (Gribov horizon in the continuum; Wightman axiom survival) 

 SL-5 / R3 (continuum ULFCB) 
 REDUCED-TO-AXIOM (via granularity screen P1) \(\to\) rolls up into CERTIFIED-IRREDUCIBLE at the gate level 
 Proven irreducible inside this framework : Bałaban rarity \(\ne\) domination, marginal \(d=4\) has no spare coercive margin; identical in kind to the Clay wall, not created by this program's choices 

 R4 ( \(a_6^\partial\) boundary coefficient) 
 OPEN/UNMADE , no sign asserted 
 Finite, in-principle-computable, genuinely UQF-3-specific (not shared with Clay); shares its GT-ladder machinery with Gap-01 

 R1 ( \(d=4\) boundary anomaly class, feeds Q² = 0) 
 AUDIT (inherited) , EXPORTED \(\to\) UQF-4 
 \(d\le3\) bulk vanishing is PROVEN (FOS Cor 7.5); \(d=4\) bulk is OPEN and flagged likely nonzero (FOS Cor 7.6) 

 R5 (nonperturbative confined-QCD IR positivity) 
 OPEN , EXPORTED \(\to\) UQF-11 
 IR certificate cannot come from UV data alone; includes the Gap-02 mass gap 

 R6 (full KK tower, all-loop positive-definite inner product) 
 OPEN , DEMOTED to conjecture 
 Prior "factorization reduce" retracted; interacting measure does not factorize base \(\times\) internal; nonseparability is load-bearing 

 R7 (above-cutoff / nonlinear graviton positivity) 
 DERIVED-GIVEN-E (linearized) / OPEN above cutoff , EXPORTED \(\to\) UQF-14 
 Needs a UV completion of gravity, blocked on UQF-9 

 R8 (OS RP \(\leftrightarrow\) BRST equivalence, interacting branch) 
 DISCLOSED-CORRECTED / OPEN (scoped) 
 Folds into R3/R6; OS RP demoted to "chosen sufficient Euclidean certificate," not the intrinsic wall itself 

 Composition theorem (local certificates \(\to\) global positive theory) 
 UNPROVEN — named theorem-debt , not assumed 
 Local \(\not\Rightarrow\) global in general; never invoked as if proved 

 7. Exact-rational cross-checks (target-blind, all independently verified)

 Quantity 
 Exact value 
 Cross-check method 
 Kind 

 \(a_6(S^2)\) (round unit) 
 \(4/315\) 
 two independent routes agree to \(\sim4\times10^{-14}\) 
 derived 

 \(a_6(S^4)\) 
 \(74/63\) 
 same calibration family 
 derived 

 \(a_6(S^6)\) (round unit) 
 \(1139/63\) 
 same calibration family 
 derived 

 \(a_6(S^6)\) (conformal) 
 \(5/63\) 
 distinct normalization, correctly kept separate 
 derived 

 \(a_4/a_2^2\) (scale-free ratio) 
 \(66/125\) 
 Levi-Civita-immune, robust across conventions 
 derived 

 \(\mathbb Z_2\) defect product-trace, \(S^2\times(S^1/\mathbb Z_2)\) 
 \((1/2)\cdot(4/315)=2/315\) 
 peels to \(\sim1\times10^{-5}\) (peel-noise); Vandermonde route \(1.3\times10^{-14}\) 
 derived 

 Twisted-circle trace \(\mathrm{Tr}(\sigma e^{-tD})\) 
 \(1.000000000000\) exactly, \(t\) -independent 
 integer powers only \(\Rightarrow\) no half-integer boundary tower 
 derived 

 \(\mathrm{tr}\,\gamma_{\rm grav}\) 
 \(67\) 
 \(=(\mathrm{tr}\,A^2+(\mathrm{tr}\,A)^2)/2\) , \(A=\mathrm{diag}(1_{12},-1)\) , \(\mathrm{tr}\,A=11\) , \(\mathrm{tr}\,A^2=13 \Rightarrow (13+121)/2=67\) 
 derived 

 \(\mathrm{tr}\,\gamma_{\rm ghost}\) 
 \(11\) 
 \(=\mathrm{tr}\,A\) 
 derived 

 Block-A graded weight 
 \(45\) 
 \(=67-2(11)\) , forced by \(D=13\) + reflection 
 derived 

 Ungraded \(\mathrm{Sym}^2\) multiplet count 
 \(91\) 
 distinct from \(67\) ; never substituted for it 
 derived 

 \(\|\mathrm{Riem}\|^2(K_6)\) 
 \(23/12\) , ratio \(23/75\) 
 frozen negative control — never \(31/147\) , never \(60\) 
 derived 

 \(K_6\) scalar \(a_6/a_0\) ("color factor \(124/315\) ") 
 \(124/315\) 
 dual-validated (spectral + algebraic) vs. the four sphere cross-checks to 13 decimals 
 derived-PENDING-independent-target-blind-reproduction — same engine that carries the R2 flag produced this; not presented as a settled clean win 

 8. Anti-claims and negative controls

 Bright-line non-claims (never printed as proved by this gate): 
- NOT "closed" — this is CERTIFIED-IRREDUCIBLE with a named open continuum leg, not a solved theorem.
- NOT a blanket "reflection positivity is derived" — only the finite-cutoff leg is derived.
- NOT "continuum positivity solved" — REDUCED-TO-AXIOM is a disposition of an open question, not a solution of it.
- NOT "UQF-3 \(\Rightarrow\) mass gap" — the scope theorem runs strictly gap-machinery \(\to\) positivity; inverting it is a named error mode.
- NOT "UQF-3 derives given-E" — Shape supplies the target spectrum; UQF-3 never certifies the spectrum, only checks positivity of the states it is handed.
- NOT "the \(d=4\) boundary anomaly vanishes" — it is uncomputed and flagged likely nonzero ; assuming vanishing would be a target-loading trap and is explicitly rejected.
- NOT "free-field textbook equivalence proves the interacting branch" — step 5's equivalence is scoped to free fields only.
- NOT "local certificates compose to a global positive theory" — the composition theorem is named theorem-debt, unproven, never assumed true.
- NOT "hashes or reproducibility validate the physics" — reproducibility of a computation is not evidence for the physical claim it computes.

 Negative controls (frozen values a correct computation must land on, and must never reproduce): 
- \(\|\mathrm{Riem}\|^2/\mathrm{Scal}^2 = 23/75\) , never \(31/147\) (withdrawn Nomizu-sign-bug artifact).
- \(\|\mathrm{Riem}\|^2 \ne 60\) (that is the unrelated round-unit \(S^6\) value).
- \(\mathrm{tr}\,\gamma_{\rm grav}=67\) must never be substituted with the ungraded count \(91\) .
- \(\mathrm{tr}[a_6] = -2.817995812\times10^{94}\ {\rm GeV}^6\) — RETRACTED , structurally ill-posed at odd \(D=13\) (half-integer heat-kernel pole, power divergence, zero in dimensional regularization); never revive, admissible at most as a labeled scale-riding consistency coefficient, never a physical positivity input.
- "Factorization reduce" for the KK tower — RETRACTED , demoted to the open conjecture R6.
- " \(-16/315\) two-route agreement" — REFUTED/FALSE , not an independent confirmation of anything.
- Curvature input \(0.2109\) used in an earlier internal \(a_6\) engine pass — DISCLOSED-CORRECTED , roughly 31% low against the true \(23/75=0.30667\) ; any magnitude computed with the old input is untrusted until re-run.

 Negative-control test explicitly passed at the root level (§2.4 above): a granularity attack (invoking P1) on the finite \(a_6^\partial\) object buys nothing — R4 survives on the open roster and is not dissolved by the continuum axiom. This demonstrates the roots are not being used promiscuously to manufacture closure; P1 dissolves only the genuinely continuum-scoped question (SL-5/R3), and is checked not to over-reach onto the finite one.

 9. Residual ledger — R1–R8, composition theorem, curvature-fix, each on its own row

 ID 
 Residual 
 Disposition 

 R1 
 \(d=4\) boundary Chern–Simons/Dai–Freed/eta-class anomaly, feeds \(Q^2=0\) 
 AUDIT (inherited) , EXPORTED \(\to\) UQF-4; \(d\le3\) PROVEN vanishing, \(d=4\) OPEN and likely nonzero 

 R2 
 Quartet/no-ghost verified only perturbatively 
 DERIVED-GIVEN-E (perturbative, retained) / OPEN (non-perturbative) 

 R3 
 Constructive positivity for the full interacting 4D base 
 REDUCED-TO-AXIOM / CERTIFIED-IRREDUCIBLE — identical to the 4D-YM Clay existence/positivity wall; no lever inside the framework 

 R4 
 \(a_6^\partial\) mixed-boundary heat-kernel coefficient on \(F\subset K_6\) 
 OPEN/UNMADE , finite, in-principle computable, shared with Gap-01; no sign asserted 

 R5 
 Nonperturbative confined-QCD IR positivity 
 OPEN , EXPORTED \(\to\) UQF-11 

 R6 
 Full KK tower + all-loop positive-definite inner product 
 OPEN , DEMOTED to conjecture 

 R7 
 Above-cutoff / nonlinear graviton positivity 
 DERIVED-GIVEN-E (linearized) / OPEN above cutoff , EXPORTED \(\to\) UQF-14 

 R8 
 OS RP \(\leftrightarrow\) BRST equivalence, interacting branch 
 DISCLOSED-CORRECTED / OPEN (scoped) 

 composition thm 
 Local certificates \(\to\) global positive theory 
 UNPROVEN — named , not assumed 

 curvature fix 
 Bianchi-consistent \(23/75\) replaces retired \(31/147\) 
 DISCLOSED-CORRECTED bookkeeping fix, not a closure 

 Three-axiom floor for the full theory (from the compositionality finding, distinct from the two-axiom atomic floor in §3): AXIOM-ANCHOR \(\cdot\) AXIOM-QUOTIENT ( \(\mathcal H_{\rm phys}=\ker Q/\mathrm{im}\,Q\) ) \(\cdot\) AXIOM-CONSTRUCTIVE-RECONSTRUCTION (full positive reconstruction across interactions, KK tower, boundaries, IR, and above-cutoff gravity). The named route to closure is the GLOBAL-ADMISSIBILITY-COMPOSITION THEOREM — theorem-debt, explicitly not an axiom substituted in its place.

 10. Endpoint line

 CERTIFIED-IRREDUCIBLE / RESOLVED +0. 

 The finite-cutoff reflection-positivity leg — \(T=e^{-aH_a}\) , \(0\le T\le1\) , \(H_a\ge0\) , Perron–Frobenius-unique vacuum, strict finite-volume gap \(\Delta(a,L)>0\) — is a real, banked DERIVED result, checked directly against Osterwalder–Seiler (1978), Lüscher (1977), Seiler (1982), and Münster (1981), with no number borrowed from elsewhere in the framework.

 The single remaining leg — continuum-uniform positivity for the fully interacting 4D base, SL-5/ULFCB \(\equiv\) R3 — is the shared Clay-class Yang–Mills constructive-positivity wall. It is proven irreducible inside this framework (Bałaban rarity \(\ne\) domination; marginal \(d=4\) Yang–Mills carries no spare coercive margin to supply uniform coercivity for free), it is named honestly as identical in kind to a wall the entire field faces, and it is dispositioned REDUCED-TO-AXIOM via the granularity screen P1: the finite-cutoff theory is declared the physically relevant regime, and the continuum limit is carried as an axiom-conditional idealization rather than a solved theorem. This is a limit on all knowledge, not a defect specific to this construction.

 The residual family — R1, R3, R4/ \(P(a_6)\) , R5, R6, R7, R8, and the composition theorem — is carried row-by-row (§9), never rolled into a single hedge, and never silently upgraded toward "closed." The scope theorem (§1, §4 step 7) is the one genuinely new banked result: reflection positivity is a proper sub-wall of the mass-gap problem, not the reverse.

 Anchors consumed: zero of \(\{M_{\rm Pl}, \alpha_i, y_t, |V_{us}|\}\) — the gate is scale-blind throughout, with no \(\sigma\) -pulls. Ceiling: serious candidate / partial unification — not validated as a completed constructive proof.