Two axioms. One compiler. A testable Standard-Model skeleton.

The even-Hamming code on five slots is the D₅ half-spinor: its dimension is the Pascal sum 1 + 5 + 10 = 16, which forces g_car = 5 uniquely. The E₈ root count is then a pure carrier trace. The current reduction state of P2 (2026-07-22, markers unchanged): the compiler's dimensionless inputs reduce to four marks (derived from P1-side topology, v216), one discrete symmetry-lift bit (flag transitivity of the four marks, V₄ → D₄, ⟺ τ = i — bare mark-transitivity is automatic on the free circle and no local jet sees the side bit, v491/v499/v506/v507/v510/v512) and π; AX.P2.01 stays the declared axiom.

The seam owns exactly one measuring device — a single scalar two-point kernel — and four theorems pin what it can do. A scalar kernel exists iff it pairs the two sheets (exactly 2 = |ℤ₂| kernels = the glue ambiguity, v110); the certified channel counts the code by itself — one neutral kernel per code state, graded (1,5,10), so the Pascal closure is two countings of one set, not a condition (v112); pair transport is minimally complete — degree ≤ 1 generates nothing, degree 2 generates every code operation (v111); and the carrier net is 16 free Majorana fermions whose tower carrier → SO(16)₁ → E8₁ never changes the field content — only the certificate grows, and the central charge is the rank of the one kernel: 5 on the carrier block, 8 on the seam hull (v113). The interior is free; the structure is the certificate. Honest residue: the premise 'the seam is the free c=8 net' is the G_net gate itself — one theorem now closes both the metric story and the carrier choice — and that premise is itself no longer free-standing: it is a fixed-point theorem whose only residual factors into the already-open A2 (net existence) and GATE.QGEO, with the irreducible core {π, v_geo} a theorem (v160–v165).

D₅ = so(10) (spinor 16) and A₃ = su(4) (the four-puncture family geometry ℙ¹∖μ₄) have the same discriminant group ℤ₄. Their discriminant-form norms are two TFPT constants that add to the E₈ root norm, so the glue closes as a lattice theorem — not a posited 248.

The Coxeter number of E₈ is h = 30 = 2·3·5 — exactly the three discrete atoms (sheet ℤ₂, families ℤ₃, carrier g_car = 5). The rank is the count of live phases of the order-30 cycle.

The fine-structure constant is the unique positive root of a parameter-free cubic built only from c₃, the abelian coefficient (Σ L + N_Φ = 41 = 10 b₁) and the exact seam generating function. Existence and uniqueness are proved; the value lands 1.9σ from CODATA-2022. The abelian coefficient is pinned three independent ways — carrier algebra 10 b₁ = g_car·2^(g_car−2)+1 = 41, the U(1) hypercharge index, and the external RGE generator PyR@TE 3, which reproduces β_g₁ = (41/10)g₁³ verbatim (v159) — so the EM input is not a free knob. The three terms reassemble as the stationarity of a U(1) determinant line (Maxwell α³ + Calderón −2c₃³α² + transport), every coefficient a named index/heat-kernel/discriminant atom (v341/v342). The one residual — the from-first-principles proof that this IS the exact ζ-regularised Quillen functional — is the tracked external target ALPHA.QUILLEN.EXACT.01 (v382), never the value. Four honest steps narrow it without closing it: a solvable 4D model reaches the a₄ heat-kernel order (v433); the matter factor b₁ is the U(1)_Y a₄ coefficient via the β = a₄ theorem, collapsing the three residuals to one [C] (the seam F-normalisation) + one [O] (v434); and a π-power test isolates the cubic α³ as the unique metric-independent (π⁰) topological rung, whose coefficient is a conditional integer Chern-Simons level (v435). A fifth step (v470) upgrades both leftovers: the α³ level equals the computed bulk Chern invariant |C| = 1 of the same collar model that realises S3 (TKNN/Avron–Seiler–Simon quantisation + Callan–Harvey inflow + the APS/Witten η=CS reading of δ log det), replacing v435's minimality assumption; and the seam F-normalisation is the affine embedding index k_Y = tr(Y²)/tr(T₃²) = 5/3 (Ginsparg 1987; (3/5)·(41/6) = 41/10 = b₁ exactly) — zero independent content, a face of SEAM.EQUIV.01. One invertible phase, two quantised responses (c₋ = 8 gravitational, C = 1 U(1)). A sixth step (v472) exhibits the bridge lemma at the finite level: the determinant line of the occupied frame over the U(1)-twist moduli of the same collar — the moduli space of flat U(1) connections, the Quillen-shaped object the target names — carries FHS curvature = 1 = the inflow level, exactly and size-independently, with clean controls and the twist-moduli integer equal to the Bloch-BZ integer (Niu–Thouless–Wu); what stays [O] is the continuum ζ-det identification on the abstract seam (= the SEAM.EQUIV.01 face). A seventh step (v484, SEAM.CONTACT.UNIT.01) unifies this target with the φ₀-puncture target: the shared 'c₃ per boundary insertion' rule (the {0,3,6} ladder here, the per-mark weight there) IS the KMS seam unit 2π = 1/(4c₃) with 1/4 = 1/|μ₄| — one bare boundary propagator orbit-averaged over the four marks — derived on the seam circle for the finite cycle sector (the bare Green function takes integer multiples of c₃·ln2 at the μ₄ separations; the log-det contact expansion is exactly graded in c₃ per insertion; the Λ prefactor 3/(4π²) = 48c₃² carries the same Ω_adm = 48 at two insertions). The two [O] targets merge into one remaining analytic step (diagonal ζ-renormalisation + multiplicity matching) — and an eighth step (v485, SEAM.CONTACT.UNIT.02) settles that step at the computable level: the renormalised diagonal vanishes EXACTLY at the KMS seam circumference (G_reg(0;ℓ) = (1/π)ln(ℓ/2π), zero iff ℓ = 2π = 1/(4c₃)), the mark determinant resums in closed form (det(I−uC) = (1−4u)(1+2u)², BFK route v151; linear term absent because Tr C = 0), and the 48/41 multiplicities are ONE state set under two response weights (flat = Ω_adm vs Y²/Ginsparg = 10b₁ = 40+1, Tr₁₆Y² = 10/3 exact). Every finite piece of the merged target is proven; the single remaining [O] is the abstract-seam ζ-det identification — a face of SEAM.EQUIV.01, the v382 typing now substantiated computationally. ALPHA.QUILLEN.EXACT.01 stays [O]; α⁻¹ stays [E].

The same α⁻¹ ≈ 137 generates the electroweak scale (divided by the carrier 5), the cosmological constant (times 2) and the Hubble scale (via the square root) — the action ladder 1 : 5 : 10 is the Pascal row of the carrier.

Every fermion mass is the same ladder: the geometric VEV times the carrier base raised to a compiler word-length, times an O(1) residue. The word-lengths are the fixed residue matrix of the compiler — not free parameters.

The word-length matrix L = R + 6·(winding) carries only compiler numbers: its trace is N_fam², its determinant is h(D₅) = 8, its principal 2-minors are (2,3,5) with product 30 = h(E₈), and its Frobenius norm is dim E₆ = 78.

The lepton amplitudes are the rationals (16/7, 4/3, 7/6) with product 2⁵/N_fam² = 32/9, and the masses are exact φ₀-powers. Applied to the down sector the lepton law provably fails — the quark c's live on the parabolic wall.

The quark mass ratios are pure integer Plücker readouts on the derived selector stratum — no transcendental solve. The absolute amplitude reduces to one overall scale v_geo (ratios + Grand Mass Volume), the same dimensionful anchor as gravity's 1/G. The remaining ℤ₃ deck choice is since derived: the integer deck pairs the Q₊=1 line with the self-conjugate character 2, so the geometric boundary deck is the sheet-twisted class and the cusp exponential is excluded (v141) — GATE.QGEO keeps only its realisation premise, with no discrete freedom left — and that premise sits at its floor: the full Möbius D₄ of the seam curve matches the integer model parity by parity (ι = T_A exactly, δι = Σ; v146). The finite rigidity is now proven exactly [E]: μ₄ has cross-ratio 2 and a faithful Möbius D₄ stabiliser, H¹(ℙ¹∖μ₄) has rank N_fam = 3, and the eigenforms ω_k carry the μ₄ characters of weights (1,2,3) = the A₃ exponents = Spec(Q₊) — so only the seam-collar realisation stays open (v168). The Sheet Diamond carrying these operators is a discrete geometry with two axes (v218): the determinant is linear along the winding axis (A₃-driven, slope 6 = |R⁺(A₃)|) and quadratic along the sheet axis with curvatures (8, 6) = (rank E₈, |R⁺(A₃)|); the anchor-Plücker coordinates lift K→C→F in two exact steps (1,8,10) then (1,8,16) — the decuple A_Λ then the full spinor generation dim S⁺; and the characteristic-polynomial discriminants of Q,K,C,F factor as q(r)²·Disc(q) with squares (1,3,4,6) = (N_Φ,N_fam,|μ₄|,|R⁺(A₃)|) and kernels (13,48,65,105). No new numbers — it organises the existing operators more strictly. Sharper still (v410): the sheet axis V = Q·diag(0,1,1) is a binary internal compiler — its powers print the carrier spine Vⁿ·1 = (2ⁿ⁻¹, 2ⁿ, 2ⁿ⁺¹−1) = (1,2,3),(2,4,7),(4,8,15),(8,16,31), and four exact bilinear families collapse the recurring integers (6,7,9,11; 13 = Δ_Q, 27 = 1ᵀRa, 55, 56 = dim 56_E₇) into one operator's iteration. The quark ratio is then a pure V-power readout c_u/c_d = (1ᵀV⁴1)/((aᵀV1)(1ᵀV²1)) = 55/117 (v411, an exact re-encoding); the unnamed Z₂-wall corner J = M(1,−2) carries χ_J = (6,3,2), aᵀJa = 30 = h(E₈), det(I+J) = 12, det(2I+J) = 40 (v412); the sheet axis encodes the atoms as difference orders Δe₁ = 3, Δ²e₂ = 4, Δ²e₃ = 8 with anchor energy 52+11t (v413); and the center C is a resolvent portal det C = 14 = dim G₂, det(I+C) = 52 = dim F₄, det(2I+C) = 120 = |R⁺(E₈)| (v414). All [E] algebra; the binary spine is forced by Spec(V) = {0,1,2}, so the Lie-dimension readings stay [C], audit-typed.

The absolute ν-mass scale is one seesaw ratio m₃ = (y_ν v)²/(2M_R) — honestly typed as one open UV input (v272), with the NO floor Σm_ν = 0.0586 eV as the cosmological kill test. New (v481, FLAV.NUSCALE.02, CANDIDATE class like v467/v468): the y_ν = 1 probe was not the carrier normalisation — one SO(10) 16 per family with the minimal Yukawa sector (10_H / PS (1,2,2)) forces y_ν = y_t at the matching scale, collapsing the (y_ν, M_R) trade-off to M_R alone. With explicit 1-loop RG (gauge/y_t/λ up, ADKLR Weinberg-operator running down) the observed m₃ = 0.0503 eV demands M_R = 9.3×10¹³ GeV — inside the compiler's own PS window [4.2×10¹³, 2.4×10¹⁵] GeV (v249) at log_c₃(M̄/M_R) = 3.15 (y_ν = 1 gives the structureless 2.58). Honesty gate: the integer rung M_R = c₃³M̄ predicts m₃ = 0.030 eV, 40% low at 1-loop, so the ladder pin is DECLINED per the anti-numerology rule — and the named decision computation is meanwhile EXECUTED in bracketed form (v482, FLAV.NUSCALE.03): the rung needs a rescue factor ×1.670 while >3σ-generous input envelopes (m_t ±3 GeV, α_s, κ-run ±10%, PS-leg β ×[0.5,1.5]) reach at most ×1.165 combined, so the unstructured rung is EXCLUDED (not an RG artifact); the only escape, a third-generation Majorana structure factor r ≈ 1.67 sitting 0.18% from g_car/N_fam = 5/3, is recorded post-hoc and declined (no forcing mechanism). That escape is now DECIDED dead (v488, FLAV.NUSCALE.04): since the 126bar is not in the E₈ hull (v247), M_R can only come from the d=5 operator (16·16bar_H)²/Λ with singlet/45 insertion channels; every ν^c channel weight {1, 1/4, 1, −1/2, 3/8} is a {2,3}-smooth rational, so no channel combination can produce 5/3 — and the unique natural 5 of the embedding, k_Y = 5/3 (Ginsparg), is a full-multiplet trace whose direction has Y(ν^c) = 0 exactly, structurally decoupled from the Majorana operator; Clebsches are generation-blind (family-space scalars), so diag(1,1,3/5) cannot arise from group theory at all. A clean negative: the rung+5/3 rescue is a numerical coincidence without mechanism, and the one-parameter window candidate stands as the honest endpoint. The candidate band m₃ ∈ [0.002, 0.115] eV brackets the observation and DESI cuts the window from below; nothing closes and the frozen record is untouched.

Tri-bimaximal gives 1/3; the charged-lepton 1–2 misalignment is the seam ε = q(A₃)φ₀ = (3/4)φ₀, and TBM geometry gives the only previously open SM angle as a conditional derivation — 0.1% from NuFIT 6.0.

At the two branch points of the anchor-block double cover the block is rank 1, with integer kernels — at the carrier point the kernel is the democratic vector itself. Rank 1 forces the kernel image onto the antisymmetric direction (−1,1,0): up and down are the deck-odd pair, and the lepton pairing vanishes — the leptons sit on the ramification (Koide is leptonic). The anchor-forced cusp conjugation T_A (with a = e₂+e₃, the conjugation-symmetric vector) realises the same deck action, and the dictionary 'Q₊ grading = A₃ discriminant grading' is now derived (G = T_A·Σ acts integrally as the B₁⊕E decomposition on the cusp basis): the sheet question carries no separate [C] — its residual coincides with the one existing Q-geometry gate.

Each maximal subalgebra of E₈ projects a TFPT module. The strongest new hit: E₆ × A₂ reads the flavor residue matrix, with E₆ reading its Frobenius norm and A₂ the three-family symmetry.

Sharper than seven separate hits: all seven 248-slices are ONE projection of just two alphabets — the anchor power sums P = (3,4,6,10) (p_n = 2+2ⁿ) and the Sheet-Diamond operator determinants (det Q,…,det F) = (3,4,8,14,20,32), which sum to 81 = N_fam⁴ (the total determinant charge of the flavour space). The K₄ edge products of P are the carrier blocks (12,18,30,24,40,60); one affine map (7,11,13)+s(3,3,3)+t(1,2,3) generates every flavour row budget; and dim E₆ = 78 = p₂(p₀+p₃). The atlas is a closed grammar over admissible invariant classes, not a shelf of trophies — exactly what the No-Free-Pattern discipline demands (v170).

An explicit, fully declared null model: the entire complexity-matched formula grammar (provably containing every scored TFPT formula) is enumerated exactly against conservative data windows over the 13 scored frozen-registry observables; Monte-Carlo pseudo-theories and negative controls (seed perturbation, data shuffle) demonstrate the test's power; all 94 500 variants of the F_U(1) equation are root-solved with exactly one CODATA hit. The result is a null-model rejection conditional on the declared grammar — never 'certainty'.

The old E₈ orbit cascade is the same even-integer spine: it starts at 60 = 2·3·10, ends at 8 = h(D₅) (the flavor selector), passes the Coxeter rung 30 = h(E₈), and the product of endpoints recovers the root count.

The carrier rank is an overdetermined E₈-closure fixed point: rank-fill (g + 3 = 8), Coxeter-match (h(D_g) = 2g − 2 = 8), and the integer-glue/norm closure whose reverse-glue quadratic has nontrivial root μ = 4. A fourth, arithmetic face (v491): given the four marks, the unique partition of 4 into three positive parts is {1,1,2}, so g_car = e₂ = 5 is a corollary of the weight typing — the residual is the typing postulate, P2 stays the declared axiom. The typing is itself hardened (v499): Deligne + Birkhoff–Grothendieck + Mehta–Seshadri stability derive integrality, positivity (h⁰ = 0) and the sum 4 from the four marks, rank 3, the cusp class and (U) — leaving the module identification plus (U) as the [C] residuals; the module-identification rest is now ONE residual with the QGEO modulus rest — both reduce to the same order-4 clock = Coxeter of W(A₃) = the cusp-class carrier datum (v503, QGEO.EMERGE.LIGHT.01: the free field alone does NOT select the square — the global det′ winner is hexagonal — the square is pinned by the clock's rigidity; markers unchanged); the clock-rigidity theorems then reduce that common residual to one alignment bit with the clock's order fermionically forced, U² = (−1)^F exactly (v506, SEAM.CLOCK.RIGIDITY.01), and the bit-origin theorems derive the deck's CLASS (a half-period translation of the seam double) — only its position stays the [C] carrier input (v507, SEAM.BIT.ORIGIN.01); the freedom theorems (v510, SEAM.BIT.FREEDOM.01) then type that position half as topology (the covering deck is free on the seam circle; the edge class is excluded), so the residual is the square-modulus datum τ = i alone; and the flag-transitivity web (v512, SEAM.TAU.FLAG.01) restates that residual as ONE discrete symmetry-lift bit — flag transitivity of the four marks (V₄ → D₄) ⟺ τ = i, with bare mark-transitivity automatic for every configuration — the current reduction state of P2: four marks (derived from P1-side topology) + one discrete symmetry-lift bit (flag transitivity ⟺ τ = i) + π, with AX.P2.01 still the declared axiom.

The seam denominator is fixed five concordant ways. The two axioms collapse to one continuous primitive (π, from Möbius/Gauss–Bonnet) plus one discrete fixed point (the E₈ closure).

The seam's boundary 2-sphere is the celestial sphere of null directions (Möb ≅ PSL(2,ℂ), the clock the order-4 Möbius map, v180). Reading the seam through the celestial-holography lens (Bittleston–Homans–Sharma's chiral algebra on ALE spaces; Costello's twistorial SDYM) yields a coherent twenty-step story — the narrative arc from the μ₄ clock to the (E₈)₁ boundary shadow — the executed work packages WP1–WP5d (both WP5d stages) plus the WP5e α, β, γ, δ₁, δ₂, ε₂ and ε₁ stages and the three back-reaction milestones M1–M3 of the research contract CELEST.SEAM.01, plus the α stage of the OS twistor bridge WOIT.OS.TWISTOR.01 (Step 20; full technical reference: the Research Contracts companion). Step 1, setup [E] (v492): the E₈ μ₄-glue grading 240 = 52+64+60+64 is INNER — a flat ℤ₄ monodromy on the A₃ ALE space ℂ²/ℤ₄ (= the A₃ singularity XY = Z⁴); the glue-equivariant sector closes (possible only because dim g_j = (60,64,60,64)) and reproduces the (E₈)₁ character as the sum of the four glue sectors; critical correction: clock² = deck (spin bridge ℤ₈, 8 = 2|μ₄| — the c₃ = 1/(8π) winding integer); that spin bookkeeping is now FORCED on the seam fermions (v506): the NS implementer of the deck satisfies U² = (−1)^F exactly (nonsplit ℤ₄, no phase choice escapes it), the quarter-shift lifts to a canonical ℤ₈ tower (V² ∝ U, V⁴ ∝ (−1)^F), and the R-sector control splits (U_R² = +1) — the forcing rests exactly on the established bounding spin structure. Step 2, deformation [E] (v493): the clock-invariant family XY = Z⁴ + a₀ carries no shape modulus (I = 12a₀, J = 0 identically ⇒ j = 1728, τ = i frozen for all a₀); the clock IS the Picard–Lefschetz/Coxeter monodromy. The residual behind that clock is now ONE alignment bit (v506: a marks-preserving root exists iff the deck is the central involution of the mark-D₄), and the bit has survived its tautology attack (v507, SEAM.BIT.ORIGIN.01): the marks are the four half-periods of the seam double and every mark-free collar deck IS a half-period translation — the deck's CLASS is derived — yet the origin does not force the alignment (core solve λ ∈ {−1, 2, 1/2}, one value per half-period; the aligned one is the CM-fixed c* = (1+τ)/2; the silver counterexample sits in-family); the bit is an equivalence tetrad (clock ⟺ τ = i with the CM-fixed half-period ⟺ deck central ⟺ harmonic deck pairs ⟺ nonsplit NS Fock lift) with a physical face — the Fidkowski–Kitaev-type extension class (central U² = (−1)^F vs edge U² = +1, split; 2 vs 0 roots) — distinguished but not derived: only the POSITION (which half-period) stays the [C] carrier input; no marker moves. And that position half is now TOPOLOGY (v510, SEAM.BIT.FREEDOM.01): the collar deck is a COVERING deck (the ℤ₂ seam of the Möbius double), hence free on the seam circle — the unique free involution is the antipode (Aut(C₁₆) = D₁₆, complete census; every reflection has two circle fixed points and its quotient is an interval, breaking the closed-circle 6π budget), NONSPLIT ⟺ FREE over all 17 involutions in exact Cl(16), and the edge/silver arrangement (whose mark circle passes through the deck poles — the restriction of the BRANCHED pillowcase involution) is excluded; harmonic + free solves exactly to b = ±i, so the alignment bit reduces from 'harmonic AND central' to the square-modulus datum τ = i alone — and the flag-transitivity web (v512, SEAM.TAU.FLAG.01) gives the modulus half its sharp discrete face: on the free circle bare mark-transitivity is automatic for every deck-invariant configuration M(δ) = {±1, ±e^(iδ)} (the pair-exchanging V₄; no local jet sees the side bit), and what selects δ = π/2 ⟺ τ = i is exactly FLAG transitivity (V₄ → D₄ symmetry lift), welded into a 12-fold exact equivalence web (odd-doublet split (2/π)ln cot(δ/2), arc-Laplacian degeneracy, the K3 indicator in closed form, ρ-twist existence, cusp-4-cycle implementability, harmonicity, j = 1728) — the carrier input reduces from a continuous modulus to ONE discrete symmetry-lift bit, its negation concretely measurable. Step 3, consistency honestly demoted [E]/[C] (v495): E₈ is on Costello's axionic list and (κ/c₃)² = 12 exactly — but the alignment format passes 8/8 across the whole list: alignment survives, selectivity does not (compatibility, not evidence). Step 4, type mismatch + boundary limit [E] (v496): the (E₈)₁ character is NOT a conformal block in the jet grading (growth n^(2/3) vs n^(1/2)), but one full μ₄ period of loop energies sums exactly to 248 — the character is a boundary-limit SHADOW, the same scaling-limit shape as SEAM.EQUIV.01/MMST. Step 5, the null ideal quantitative [E] (v497): stabilisation theorem w = n+1; 27000 = 30³ = h∨³ DERIVED (Freudenthal/Weyl/peeling); quotient 31124−27000 = 4124 = the independent μ₄ sector sum (two routes, one number); SO(16)₁ negative control (four blocks, 5304 ≠ 14³). Step 6, the deleting object as an operator [E] (v498): |s⟩ = (E^θ_{−1})²|0⟩ explicit, J^a_1|s⟩ = 0 for all 248 generators (case tally 190/57/1), level dial 2(1−k) (only k = 1 deletes), clock phase −1; D₈ contrast: the singular-vector mechanism is level-1 generic, the one-block closure is the E₈/μ₄-specific part. Step 7, the limit state [E]/[C] (v500): the quasi-free family ω_w stabilises exactly at the WP5a threshold and its limit carries the null ideal in its GNS kernel — the complete 9361-block exact level-2 Gram has rank 4124 with kernel 27000 = V(2θ) weight by weight, |s⟩ IS the GNS zero vector, and a CCR obstruction shows the family is NECESSARY. Step 8, the two-interval index [E]/[C] (v501): the fermionic two-interval MI is extensive (μ = 1 reference) while the orbifold prescription pays exactly one classical bit (the ln 2 plateau at machine precision), so μ_gauged = 4 = the v490 parity census, and the condensation arithmetic 16 → 4 → 1 (KLM/Longo–Rehren, θ_v = 1 at ν = 16) closes with μ = 1 — the preregistered kill does not fire. Step 9, the prefactor and the level pinned on the CFT side [E]/[C] (v502): the q^(−1/3) prefactor of E₄/η⁸ is exact μ₄ vacuum-energy bookkeeping — the clock is INNER, so the twist is θ = 0⁸ in every sector (a SHIFT orbifold, not a rotation orbifold) and each sector carries −c/24 = −1/3 at c = 8; the sector weights (0,1,1,1) ARE Casimir energies (spectral flow, and exactly via the 16-Majorana seam carrier: 5/8, 3/8, 1); k = 1 is forced three independent ways (current condition h(J) = k; conformal embedding 47(k−1)(k+266/47) = 0; central charge 240(k−1) = 0) plus the Step-6 singular-vector dial (31124 − 27000 = 4124 at k = 1 only); honest sharpening: glue-h integrality holds at EVERY level and fixes nothing — the naive route is retired; the deck rotation reading fails (3/16 ≠ 3/8). Step 10, the KLM completion on the lattice [E]/[C] (v504): complete rationality (Kawahigashi–Longo–Müger) = split + strong additivity + finite μ-index, and with Step 8 supplying the index the two remaining legs are witnessed for the same orbifold prescription — strong additivity is algebraically EXACT with the shared boundary Majorana (Even(A) ∨ Even(B) = Even(A∪B), GF2 spans full, rank 32/32; disjoint exactly index 2 = the Step-8 ln 2 bit, localised at the split point); entropically the touching defect stays BOUNDED < ln 2 with the Ising ¼-exponent approach (p = 0.2444) while the preregistered U(1)/Dirac control DIVERGES (Klich–Levitov pinned): bounded-vs-divergent is the lattice discriminator (bounded ⟺ finite index, Longo–Xu); the split property is quantitative (coupling σ_k at the elliptic-nome rate πK(1−x)/K(x) to 1.3–2.0%) and the orbifold INHERITS the same ladder exactly (C → −C; the Λ²C compound — Longo heredity); Pimsner–Popa E(a) − a/2 = PaP/2 holds identically (λ = 1/2 = 1/[F:F_even] with exact integer attainment, λ_E4 = 1/4 = 1/μ, exp(Δ∞) = 2 = 1/λ_PP — two independent index routes). All three KLM ingredients now carry lattice witnesses; the continuum uplift is honestly fenced — 'finite-group orbifolds of completely rational nets are completely rational' is Xu's theorem, cited, not claimed. Step 11, the equivariant anomaly ledger on twistor space [E]/[C] (v505): the twistor side of the uplift pushed as far as exact arithmetic reaches — the Atiyah–Bott/Lefschetz fixed-point skeleton of the one-loop box anomaly on ℂ²/ℤ₄ is exact (denominators (2,4,2) with Dedekind sum 5/4 = (|ℤ₄|²−1)/12, equivariant characters (248,0,−8,0) by two routes, invariant average 60 = the carrier); the honest sharpening: only the INVARIANT sector is Okubo-quadratic (36⟨x,x⟩², 36 = λ̃²_e8), the twisted sectors are not, and the AB-weighted sum cancels the D₅ quartic exactly while leaving the RIGID residual 32·T₃ — twisted-sector closed strings must carry the rest; the index bridge is the centrepiece: f(m) = ch₂(T_m) exactly (fixed-point ledger = McKay/Kronheimer intersection ledger), glue defect −78 by both routes; the level dials say k = 1 GEOMETRICALLY (lattice current count 240/0/0/0, embedding residual (0,360,814,1362), one scale ⇒ one level); and an honest refutation: a₀ fills the BSS GRAVITON slot O(2), NOT the axion slot O(−2) (weight mismatch 4 = |μ₄|) — 'the theory brings its own GS axion as a₀' is false; instead the three H²(ALE) classes carry exactly the three twisted-sector Coxeter characters {i,−1,−i}, and the bulk axion must come from the O(−2) tower field itself; the preregistered kill ('inflow demands a level ≠ 1') does NOT fire on the equivariant skeleton. Step 12, the exchange no-go and the contact-term criterion [E]/[C] (v508): could the three sphere axions of Step 11 cancel the rigid 32·T₃ residual by Green–Schwarz-type EXCHANGE with sector-compatible quadratic vertices? No — and the refusal is a theorem, not a fit: the W(D₅)×W(A₃)-invariant vertex space on the glue Cartan is computed exactly (quadratics = span{s₅, s₃}, dim 2; quartics = span{P₁,P₂,P₃,T₅,T₃}, dim 5 — complete by Weyl nullspace arithmetic), and the PRODUCT THEOREM kills the mechanism wholesale (any two invariant quadratics multiply into span{P₁,P₂,P₃} — zero T₅/T₃ content, so every exchange image is annihilated by Φ_T3 while Φ_T3(A_fix) = 32 ≠ 0); the strict two-index ℤ₄ rule collapses the sphere couplings entirely (only m = 0 survives), the twist-insertion channels give a rank-2 exchange matrix with rank([M | A_fix]) = 3 and the annihilator certificate (Φ_T5, Φ_T3, Φ_P)(A_fix) = (0, 32, 72) — two independent obstructions, the second driven by the side discovery K⁽⁰⁾ = −15·K⁽²⁾ (the even-sector quadratics are parallel); naturalness DISSOLVES (ch₂-natural and Atiyah–Bott-weight couplings both certify (0,0) against the required (32,72) — scale- and coupling-independent); controls: SO(16) has NO sphere partners and uncancelled T₅ (E₈ doubly special), D₈ has no T₃ structure at all; the slot bijection of Step 11 is untouched and the preregistered level-kill still does not fire — what dies is the exchange REALISATION of the pairing; the 32·T₃ burden moves to the twisted BCOV contact terms with a BINARY criterion: the one-loop coefficient on ℂ²/ℤ₄ must come out exactly (9,−30,−15,0,32), computed, not fitted. Step 13, one level is a theorem and the sector counter pins k = 1 [E]/[C] (v509): the Costello–Paquette–Sharma level-from-flux mechanism (Burns holography: 'level = flux quantum × Dynkin index') executed on the lockstep spheres — the CPS skeleton is replicated exactly (S³ period (2πi)²N, exceptional-sphere Kähler flux 2πN, boundary level magnitude 2N with T(vector) = 1), and the geometric side is PINNED, not postulated: the Step-11 ch₂ ledger + integrality + unimodularity + effectivity force the pairing matrix c₁(T_m).[Σ_i] = δ_mi by complete enumeration (48 unimodular solutions, exactly 2 effective: identity + diagram flip), so the glue fluxes are F_i = (64,60,64) = dim g_i with exactly ONE quantum per charged root current; 'level = total open-string flux' is KILLED by the lockstep test itself (three unequal numbers — F_i is the flavour multiplicity), while 'level = flux per adjoint current' gives (1,1,1), anchored by the lattice current count (240,0,0,0) and the embedding index 1; the ord-4-vs-level-1 tension resolves (the clock order counts FRACTIONAL flux sectors: μ = 16 = ord², Lagrangian μ₄ glue, KLM 16 → 1) and a new dial pins the value: #primaries((E₈)_k) = (1,3,5,10,15,27) for k = 1..6 — exactly ONE sector iff k = 1; the centrepiece is the LOCKSTEP THEOREM, lifting Step 11's 'one scale ⇒ one level' from heuristic to theorem (clock invariance forces one μ₄ orbit of branch points, |Π_j| = √2·t equal on all three spheres, det(A−1) = −4: no invariant flux vector — k₁ = k₂ = k₃ IS a theorem of clock invariance), with a new falsifier (the clock-forbidden family Z⁴ − Z: zero lockstep chains over all 24 orderings) and an honest limit (uniform doubling keeps the lockstep — equality, not the value); controls: k = 2 dies on the closure dials while the lockstep dial honestly does not fire, SO(16) leaves two spheres flux-dark, A₂ admits no Lagrangian glue at all (μ = 12 not a square); the CPS dictionary on PT/ℤ₄ itself stays [C] (their geometry is the ℂ² blow-up, not the A₃ ALE), the type-I B-model back-reaction stays [O]. Step 14, the full-tensor ledger: the collapse is real, and one cubic door opens [E]/[C] (v511): Step 12's collapse was a CARTAN-restricted statement — the named escape route (δ₂) recomputes the ledger on the full adjoint tensor structure of e₈, with two exact halves. The collapse is CONFIRMED: g₀ = d₅ ⊕ a₃ is semisimple with no u(1) factor (class-0 roots 40+12, root span rank 8), the sectors factorise into minuscule Weyl orbits g₁ = (16_s, 4), g₂ = (10, 6), g₃ = (g₁)*, and the INNERNESS THEOREM decides everything wholesale (the grading element h lies in the Cartan OF g₀, so every g₀-invariant tensor carries total charge 0 mod 4): bilinear invariants exist only at j′ = −j (dims 2/1/1/1) and ALL 15 non-neutral trilinear sector triples have Hom = 0 — the sphere axions have no invariant partner operator at any arity ≤ 3, full-tensorially. AND one cubic door opens: the unique totally symmetric trilinear on all of e₈ — the su(4) d-symbol (so(10) has no cubic Casimir, T₅ stays protected) — carries an exchange quartic Q_dd = (1/60)(T₃ − P₃/4) (two independent routes, Killing propagator 2h∨ = 60 from the roots) with Φ_T3(Q_dd) = 1/60 ≠ 0: Step 12's master kill covers quadratic vertices only. The pairing remains obstructed in the charge reading (M = [E₁₃, E₂₂, E₀₀, Q_dd] rank 3 vs augmented 4) but with the genuinely WEAKER certificate {Φ_T5, ψ = Φ_P − Φ_T3/4}, ψ(A_fix) = 72 − 8 = 64; relaxed, it becomes exactly solvable: A_fix = −u + 8v + 2w + 1920·Q_dd. Number fences typed [C] only: 64 = dim g₁ = 2⁶, 1920 = |W(D₅)| = 8·240 — and the 1920 fence has since been stress-tested and FAILS as a derivation claim (v513, CELEST.DTERM.NONDERIV.01): the |W(D₅)| reading is look-elsewhere-loaded (11/924 catalog expressions hit 1920, vs 8/924 for the control target 1800) and convention-contingent (only 1 of 5 normalisations produces 1920; the charge-legal flux pairings give 2048/1800 and miss; the twisted cubic index 32i clashes in class provenance with the true quartic 32) — the convention-stable [E] core is the factorisation c_d = Φ_T3(A_fix)×2h∨(E₈) = 32×60, the fence stays [C], the physical generation of the 32 stays [O]. Controls: SO(16)/D₈ has no symmetric cubic, no odd sectors and no A₃ block (E₈/μ₄ doubly special); false g₀/sector assignments inflate or kill the tables. The burden on the δ₁ contact terms is thereby SHARPENED (they must supply either the ψ = 64 slice or the selection-rule relaxation — a burden since DISCHARGED by Step 17: the declared KS measure supplies the ψ = 64 slice exactly, so the cubic d-channel is not needed) — and the δ₁ exploration run itself stands UNDECIDED: the strict holomorphic reading is refuted at the ℤ₂/Eguchi–Hanson anchor (a method boundary, not a kill); the contact term is the MODULAR COMPLETION of the Atiyah–Bott data, a Harvey–Moore-type τ-integral with the forced leading (T₅,T₃) ratio 4:3. Step 15, the bulk axion is a construction and λ̃ = 6 is pinned three ways [E]/[C] (v514): Step 11 refuted 'a₀ is the GS axion' and left the bulk axion to the O(−2) tower field as a slot — the ε₁ stage now BUILDS it. On PT′ = Tot(O(1)⊕O(1) → P¹) the pushforward ledger π₊O(−2) closes the Penrose accounting exactly ((d+1)² classes per degree = the exact wave-operator nullspaces, d ≤ 6) and EQUIVARIANTLY: the clock acts on the fibre coordinates as (μ₁,μ₂) ↦ (iμ₁,i⁻¹μ₂), the incidence relation forces the column weights (+1,+1,−1,−1), and the bookkeeping closes block by block for all four characters; the character series are P₀ = 1 + 3t² + 15t⁴ + 21t⁶ + 45t⁸ + …, P₁ = P₃ = 2t + 8t³ + 18t⁵ + …, P₂ = 6t² + 10t⁴ + 28t⁶ + …, the d = 0 slot has multiplicity (1,0,0,0) — THE BULK AXION SURVIVES THE PROJECTION — and the invariant fibre ring is the hypersurface (1−t⁸)/((1−t⁴)²(1−t²)): Z ∈ O(2), X, Y ∈ O(4), one relation in degree 8 = the a₀ weight (the Step-1/Step-2 geometry re-emerging from the slot); the Step-11 bijection sharpens per character (twisted minimal content (2t, 6t², 2t) = the Coxeter eigenvalues, no degree-0 mode, det(A−1) = −4) and the graviton control separates a₀ cohomologically (the O(+2) slot starts invariant only at fibre degree 4 with multiplicity 3 = {X, Z², Y}). The Green–Schwarz residue is pinned THREE independent ways: Okubo (Tr_adj X⁴ = (6⟨x,x⟩)² exactly on the 240 glue roots, 36 = h∨ + 6, λ̃ = 6 = |ℤ₂|·N_fam), the measure chain on PT/ℤ₄ (the quotient factor μ = 1/4 enters vertex²/propagator/anomaly and cancels EXACTLY; wrong bookings quantified and excluded: anomaly-only ⇒ λ̃ = 3, missing propagator renormalisation ⇒ λ̃ = 12), and the flux side (minimal CPS quantum N = 1, a single exchange channel exactly at k = 1, (κ/c₃)² = 12). The first honest back-reaction step follows in Gibbons–Hawking form: clock-invariant four-centre configurations form exactly two branches (four axis points = pure resolution, one free μ₄ orbit = pure deformation), the free orbit re-derives the Step-2 family (Π_p(Z − iᵖz₀) = Z⁴ − z₀⁴, a₀ = −z₀⁴, monodromy = one clock step), the charge-4 GH point is exactly the S³/ℤ₄ seam boundary, the periods Π_j = 4πt₀(i−1)(1, i, i²) confirm the Step-13 lockstep FROM GEOMETRY (Π·A = iΠ), the source ledger carries charge 4 = |μ₄|, and the honest sharpening: the Ricci-flat ALE has asymptotic log coefficient EXACTLY ZERO (the CPS log is an exceptional-locus statement; Burns contrast det g ≠ 1), with the multipole selection rule m ≡ 0 mod 4 storing −a₀ in the first symmetry-breaking (4,±4) harmonic. Controls: so₈ (λ̃² = 12 irrational — perfect squares in the Deligne series only {9, 36} = {sl₃, e₈}), diag(i,i) (Veronese cone, no hypersurface), k = 2 (three exchange channels). Honest fences: conditional on Costello's flat-PT matching [C]; the QUANTISED BCOV coefficient on PT/ℤ₄ and the twisted channels (32·T₃) stay [O] — preregistered as the M1–M3 back-reaction milestones (the A₃ Ω_N, the twisted KS measure, the a₀ uplift), each with success and kill criteria; all three are executed in Steps 16–18. Step 16, the back-reacted Ω_N: closed form, integral periods, forced charge 4 [E]/[C] (v515): the M1 milestone preregistered in Step 15 is now EXECUTED, with the preregistered SUCCESS criterion met and neither KILL fired. The residue 2-form is DERIVED, not assumed (on F = XY − P(Z) all three ambient representatives satisfy dF∧ω₂ = −dX∧dY∧dZ and pull back to ω₂ = dX∧dZ/X; at a₀ = 0 the orbifold-cover pullback is ω₂ = 4·dz₁∧dz₂ = |μ₄| × flat form — the residue normalisation itself CARRIES the source charge 4 — and the clock multiplies ω₂ by i, the Step-1 det = i replicated); the four O(2) centre sections q_p(λ) = iᵖt₀ − i⁻ᵖt₀λ² close the twistor family in one line, XY = Z⁴ + 4t₀²λ²Z² − t₀⁴(1−λ⁴)² (e₁ = e₃ = 0 identically; the seam fibre λ = 0 is exactly the Step-2 family, the a₂ ∈ O(4) slot opens only off-seam), with the CY-compatible clock lift γ: (Z,λ) ↦ (iZ,−iλ), γ⁴ = 1, Ω → +Ω; the periods reduce generically (∫ω₂ = 2πi(q_{j+1} − q_j) for arbitrary profile and path), giving the seam-fibre lockstep vector 2πi·t₀(i−1)(1, i, i²) with the exact covariance Π_{j+1}(−iλ) = iΠ_j(λ) for ALL λ, and the 12 collision nodes — honest conifold points (Hessian det 512t₀⁶, quadric rank 4) — sit exactly on the 8 eighth roots of unity (8 = 2|μ₄|, the ℤ₈ bridge), in clock orbits 4+4+4; the back-reacted 3-form is CLOSED FORM, Ω_N = Ω₀ + Σ N_p K_p with the CPS/Bochner–Martinelli kernel on the four centre twistor lines (∫_{S³}K = (2πi)² exactly, dK = 0 off-source, scale degree 0); the undeformed Ω₀ carries ZERO quantised 3-flux (all flux is sourced), every 3-cycle period is (2πi)²-integral, the flux vector N(1,1,1,1) is forced uniform TWICE OVER (clock orbit + the K₄ connectivity of the four lines), and the lens fundamental domain FORCES N ∈ 4ℤ — the source charge 4 = |μ₄| from quotient large-gauge quantisation, minimal invariant charge ↔ the CPS quantum N = 1 ↔ k = 1 (Step 13). The honest fence has teeth: on the clock-forbidden family Z⁴ − Z the (2πi)² quantisation ALSO holds — integrality alone is NOT the discriminator; the discriminator is the lockstep phase/modulus structure (0/24 orderings vs 8/24, node support on twelfth roots instead of ℤ₈) together with the clock forcing of equal fluxes. Anchors and controls: EH/ℤ₂ forces charge 2 = |ℤ₂| from the same machinery, the CPS flat patch allows every integer N, diag(i,i) collapses at step zero (Ω₀ → −Ω₀), fractional charges N = 1, 2, 3 excluded. Global kernel patching, the Hitchin small resolution and the CPS brane dictionary stay [C]; M2–M3 are executed in Steps 17–18, and the quantised BCOV coefficient stays [O] only in its measure question (the δ₁ chain itself is decided in Step 19). Step 17, the twisted KS measure: every sector the same Okubo square [E]/[C] (v516): the M2 milestone preregistered in Step 15 is now EXECUTED, with the preregistered SUCCESS wording met ON THE DECLARED COMPLETION MEASURE and the KILL (a leftover independent quartic in ANY sector) not fired. The measure ansatz is declared before computing: M2 declares the completion contact term contact_j = (Q⁽⁰⁾ − Q⁽ʲ⁾)/det_j — in the channel with the gʲ zero-mode normalisation 1/det_j the twisted-sector LOOP contributes the UNPHASED sector trace, of which the Atiyah–Bott skeleton kept only the phase-weighted insertion part. Everything downstream is exact arithmetic with no dial to turn: the completion-weight identity w_m = Σ_j(1 − i^{jm})/det_j = (0, 3/2, 2, 3/2) = 4h_m = |μ₄|h_m = −4·ch₂(T_m) — the three sphere axions pair through their OWN McKay ch₂ charges, no free scale, no fit; the locks are parameter-free (T₅ = 0 for ANY scale, ratio 4:3 = the δ₁-forced leading ratio REPRODUCED, the T₃ budget forces c = 4 = |μ₄| uniquely); every twisted channel becomes the SAME perfect Okubo square 36⟨x,x⟩²/det_j (T₅ = T₃ = 0 in every sector), the total 45⟨x,x⟩² = (5/4)×36 = Dedekind × Okubo is exactly the unique quartic-free weighting of the Step-11 rigidity theorem; both Step-12 certificates are killed (Φ_T3: 32 → 0, Φ_P: 72 → 0) and the Step-14 slice ψ = 64 is SUPPLIED EXACTLY — the cubic d-channel is not needed (c_d free = 0); controls: wrong scale leaves T₃ = 32 − 8c, the shuffle breaks T₅ and T₃, SO(16) keeps T₅ = 20/T₃ = −40 (the KILL FIRES there: E₈ doubly special), diag(i,i) degenerates to the ℤ₂ target, the ℤ₂/EH anchor passes at scale 2 = |ℤ₂|. The mandatory fence: the completion reading (loop = unphased sector trace with the same zero-mode normalisation) is DECLARED, supported by the δ₁ modular-completion finding, NOT derived from the BCOV integral — the δ₁ chain has since been decided in Step 19 (the DERIVED chiral measure fails all three testers), so the open face is the declared-vs-derived measure question [O]. Step 18, the a₀ uplift: four coupled centre-count scales [E]/[C] (v517): the M3 milestone preregistered in Step 15 is now EXECUTED, with the preregistered SUCCESS wording ('a log-type correction whose coefficient is tied to the source charge 4 = |μ₄|') met and the KILL (decoupling from the centre count) not fired. The uplift object is the generalized-Legendre-transform kernel χ = log P₄ — the log of the Step-16 family polynomial (the GLT dictionary of Lindström–Roček / Ivanov–Roček, typed [C]); the bridge is exact: the O(2) section is a NULL coordinate (any kernel is harmonic) and the residue identity ∂_x[transform(log η_p)] = 1/r_p matches the V-ledger exactly (flux −4π per centre, source charge 4 = |μ₄|). The correction arrives on FOUR COUPLED CENTRE-COUNT SCALES: (i) the asymptotic kernel log χ = 4·log η + a₂/η² + (a₀ − a₂²/2)/η⁴ + … with coefficient 4 = |μ₄|, the seam-fibre first correction EXACTLY a₀/η⁴ = the (4,±4) multipole (exact m-grading, no log×power terms); (ii) the GLT tower p_{4k} = 4(−a₀)^k with the n ≡ 0 mod 4 selection rule; (iii) the exceptional-locus log χ(0) = log a₀ = 4·log t₀ + i(4φ₀ + π) (response 1 at the locus vs power law at infinity); (iv) the period response d log Π_j/d log a₀ = 1/4 = 1/|μ₄| uniformly, integrating to the monodromy i = ONE Coxeter clock step (the Step-2 monodromy reproduced from perturbation theory), with a₀-rigid ℤ₈ node support and topological (2πi)² fluxes. Controls: the (4,0) multipole is clock-invariant and breaks nothing; the ℤ₂/EH analogue reads 2 = |ℤ₂| on EVERY dial; k = 3, 5 orbits move the coefficient with the centre count (O(6)/O(10) slots); the forbidden family fails both dials (p₃ = 3, e₄ = 0). What stays [O]: the full nonlinear Kähler potential of the resolved A₃ ALE. Step 19, the derived measure decides — and disagrees with the declared one [E]/[C] (v518): the δ₁ question that Step 17 left open (DERIVE the completion reading from the Harvey–Moore/BCOV τ-integral) is executed as the consolidated δ₁b/δ₁c/δ₁d chain: instead of declaring or scanning a measure, the measure is SOLVED FOR from blockwise SL(2,ℤ) covariance of the dressed 16-component Weil system. Four exact results: (i) the completion closes — the discriminant module of D₅⊕A₃ is ℤ₄×ℤ₄ with q = (5x²+3y²)/8, Gauss sums 2ζ₈⁵×2ζ₈³ = 4 = √16 (signature 0 mod 8 = rank E₈), the Weil relations hold exactly, the diagonal and anti-diagonal Lagrangians are the two E₈ gluings, and the 16-vector theta S-covariance drops the naive 4-character-rule residual from 2.91 = O(1) to ~10⁻³⁹; (ii) the obstruction is a finite μ₄ CHARACTER, identified — the SL(2,ℤ) relation defects are (1,1,1) exactly on all 15 sector pairs (a character of the orbit stabilisers Γ₁(4)/Γ₀(2)±, not a genuine 2-cocycle), on Γ₁(4) explicitly λ(γ) = i^(2B+C/4); (iii) the cancellation exists and is the TWISTED FIBRE BLOCK — G[a,b] = f₁f₃ is an exact identity, the T-fix mechanism is one line of exact phase arithmetic ((−1)·e(−1/6) = e(1/3) = χ₄(T)), and the exact cancellation table reads: bare → order 4, the three sphere axions f₁f₂f₃ → 4, f₂ → 6, f₁f₃ → 1 — only the twistor-fibre content cancels the μ₄ system, with the strict solutions exactly χ₄ (dims (3,3)) and χ₁₀ (dims (1,1)), certified on the dressed functions; (iv) ALL THREE preregistered testers fail — under both derived solutions the integral misses T₅ = 0 (fractions 0.5/0.83), misses the forced 4:3 leading ratio (W₁₃ = 0) and misses −A_fix/the ψ = −64N slice (spreads 1.05/1.47), and the honest (N₁,N₂) orbit rebalance rescues nothing in the positive cone — a genuine KILL on the derived surface. Controls: the ℤ₂/Eguchi–Hanson anchor (residual order 2, cancelled to 1 by the same mechanism), SO(16) (the order-4 supply structurally absent), wrong form/wrong signature (break exactly). Fences [C]: the f₁f₃ = KS-weight identification (the deck weights (1,3) are the twistor fibre weights forced by the Step-15 incidence ledger — an exact match, NOT a complete BCOV derivation) and the kernel-family convention. TENSION, stated honestly: the DECLARED completion reading (Step 17) delivers the ψ = 64 slice and cancels the 32·T₃; the DERIVED chiral measure (this step) fails all three testers — both are exact; the sharp open question is which of the two is the true BCOV measure. Neither result is hidden behind the other. What remains (WP5e proper [O]): the GLOBAL BCOV/Kodaira–Spencer quantisation on PT/ℤ₄ (the partition function derived from the twistor side — the CFT-side dials and the equivariant skeleton are now both pinned; the exchange sub-branch is closed by Step 12, the full-tensor ledger executed by Step 14, the level-from-flux dial executed by Step 13, the bulk-axion slot built by Step 15, all three back-reaction milestones it preregistered are executed by Steps 16–18 — the Step-15 fence M1–M3 is FULLY WORKED OFF — and the δ₁ chain is DECIDED by Step 19 (kill under the derived measure); the named remaining target narrows to the MEASURE QUESTION: the declared completion reading (Step 17, delivering the ψ = 64 slice) and the derived chiral measure (Step 19, failing all three testers) are both exact and disagree — the global quantisation must decide which of the two it produces; the cubic d-channel stays not needed and its physical-justification question (c_d = 32×60; the 1920 = |W(D₅)| reading look-elsewhere-loaded and convention-contingent, v513) stays dissolved (Step 19 does not revive it)), together with the continuum uplift of the Step-10 lattice witnesses (Xu's theorem for the abstract statement; the concrete seam quotient net and the condensed (E₈)₁ net are Costello–Li territory). The route does NOT close SEAM.EQUIV.01 — it IS the keystone's second, quantitative route, named in its own right as SEAM.EQUIV.TWISTOR.01 ([O]; the conditional MMST route is SEAM.EQUIV.MMST.01, closed modulo cited theorems; the parent closes if either route closes and stays [O] as an unconditional claim) — with the ideal fixed, the operator explicit, the limit state constructed, the index chain measured, the KLM triple witnessed, both faces of the inflow anchored, the collapse confirmed full-tensorially with one cubic door open, one level a theorem with its value pinned by the sector counter, the bulk-axion slot a construction with λ̃ = 6 triply pinned, the back-reacted Ω_N closed-form with (2πi)²-integral lockstep periods and the source charge 4 = |μ₄| forced, the twisted KS measure (on the declared completion reading) landing on the unique quartic-free weighting, the a₀ uplift coupled to the centre count on four scales, and the δ₁ chain decided — kill under the derived measure, the declared/derived tension stated; the global quantisation stays open. Step 20, the real structure exists — and free reflection positivity picks the same family [E]/[O] (v519): the α stage of the OS twistor bridge WOIT.OS.TWISTOR.01 is executed (WOIT.THETA.FREE.01). The classification is complete: exactly TWO families of anti-linear structures on ℂ² normalise the clock ρ = diag(i,1) — family D (z ↦ μz̄, |μ| = 1) satisfies ΘρΘ = ρ⁻¹ EXACTLY with Θ² = +1 (−1 is IMPOSSIBLE: M·M̄ = diag(|μ|², 1) — the clock-inverting family is Kramers-free), inverts the deck and reflects the seam circle with two cut points; family A (z ↦ μ/z̄) centralises the clock projectively and NEVER inverts it, Θ² = ±1 per μ. The role separation is sharp: family A DEFINES the euclidean section — Woit's ρ_tw (ρ_tw² = −1, no real points) is replicated exactly on ℂ⁴ and is family A globally — while family D REFLECTS it (the OS conjugation; σ_std with real points ℝP³). Mark compatibility pins μ ∈ μ₄ (a 4-element torsor, ρΘ_μρ⁻¹ = Θ_{−μ}: two clock orbits); the ℤ₈ spin plane has no phase leaks; in exact Cl(16) the Fock implementer Θ_Fock = U_r∘K has Θ_Fock² = 2⁷ > 0 (normalised +1), inverts the Fock clock tower (V ↦ 4096·V⁻¹, exact scalar) and normalises the deck — while the DECK-induced candidate has Θ_t² = 256·γ₁⋯γ₁₆ = (−1)^F: the v510 split/nonsplit dichotomy IS the Θ² = +1 vs (−1)^F dichotomy, so the deck does NOT furnish the OS Θ — the seam-circle REFLECTION does. Free reflection positivity then holds for exactly that pinned Θ: on the bond cut (marks at the bond midpoints of the 16-Majorana circle) the one-particle Gram is positive definite ((8,0,0), min eigenvalue 1.888e-3 at 40 digits), the even deg ≤ 2 sector is (29,0,0), and at N = 8 the COMPLETE half-sided algebra is RP with no degree truncation; the twist η = +i is forced (η = 1 non-Hermitian); the cut THROUGH sites fails exactly (det = 0, inertia (3,3,1) — a lattice-placement artifact, the continuum Cauchy–Stieltjes control is strictly positive), and the clock-centralising family-A structure fails RP STRUCTURALLY ((4,4,0)): RP and ΘρΘ = ρ⁻¹ select the SAME family. Bonus: the anti-chiral state flips the odd sector to negative definite — an exact free-level shadow of kill test 3. Kill test 1 of the contract does NOT fire at the free/equivariant level and stays live on the interacting algebra; the interacting algebra, gauge-fixed RP, OS reconstruction, chirality and μ₄ incidence remain open contract work (the β/γ milestones named in the contract). No marker moves. THE WOIT BRIDGE (chiral Wick rotation and the real structure, [O]): one EXTERNAL programme is close enough in shape to deserve a named paragraph — named precisely so that proximity is not mistaken for confirmation. Woit's Euclidean Twistor Unification (arXiv:2104.05099) observes that the usual Wick rotation is problematic for purely chiral theories and proposes the counter-move: formulate the theory Euclidean-holomorphically on projective twistor space PT and reconstruct the Lorentzian theory from a conjugation structure / boundary values. TFPT's route has, independently, assembled the ingredients such a reconstruction needs — reflection positivity of the seam collar (v379), the OS reconstruction step (the cited AMT/OS selector in FORM.SEAM.MMST.01), the order-4 clock ρ = diag(i,1) (v492), and the seam reflection (Step 20 makes the referent precise: the seam-circle REFLECTION, not the deck/covering involution — the deck, whose freeness is topology v510, carries the (−1)^F Kramers class instead). What is MISSING is stated exactly: a global anti-linear map Θ: A(PT/Γ) → A(PT/Γ) with Θ² = 1 and ΘρΘ = ρ⁻¹ on the interacting open+closed twistorial algebra, together with reflection positivity of the interacting BCOV+SDYM functional — the two inputs of the new central contract WOIT.OS.TWISTOR.01 in the Research Contracts companion. The α-stage status (Step 20, v519): the missing global object now exists at the FREE level — Θ with Θ² = +1 and ΘρΘ = ρ⁻¹ on all four levels (sphere, ℂ⁴, ℤ₈ spin, Cl(16) Fock), with free RP on the seam system selecting the same family; Woit's two inequivalent real structures (ρ_tw and σ_std) are exactly the two families of that classification, in DIFFERENT contract slots. The interacting statement stays [O] — kill test 1 is discharged only at the free level, and until Θ + RP exist on A_hol, the Woit connection remains a strong analogy, not a mathematical integration; nothing in this paragraph is evidence for either programme. No marker moves.

From the closed baryon fraction Ω_b = (4π − 1)β_rad, the asymmetry follows as a cosmological readout. Leptogenesis is operationalised as a falsifiable interface (v169): fed by TFPT's normal-ordered neutrino spectrum and δ_CP = 240°, the thermal estimate η_B ~ 0.96×10⁻²·ε₁·κ_f brackets the observed 6.1×10⁻¹⁰ over M₁ ∈ [3×10⁹, 3×10¹⁰] GeV (a canonical M₁ = 10¹⁰ GeV gives 6.0×10⁻¹⁰, untuned). But M₁ and the washout are scenario inputs, so η_B stays [C]: if a precise Boltzmann solve excluded the window the route falls, not the theory. The cleanest scenario (v212) shares the decuple A_Λ = 10 = |E(K₅)| across both Boltzmann inputs (M₁ ≈ 8.65×10⁹ GeV, m̃₁ = m₃/A_Λ ≈ 5 meV) with no hidden seesaw scale — a sharper [C] route that cuts the free inputs from two to one, not to zero. The full BDP Boltzmann ODE solve confirms the route at the frozen M₁ (integrated κ_f = 0.092 ⇒ η_B = 6.5×10⁻¹⁰ = 1.07× observed, no free M_R dial), so η_B is a consistent [C] downstream readout, not a derivation (the flavored density-matrix solve is the next refinement).

The seam UV is the free chiral c=8 fixed point, so the one marginal SM scalar coupling vanishes there: λ(M_seam) = 0 and β_λ(M_seam) = 0 — the Shaposhnikov–Wetterich double criticality, here derived from the free seam, not assumed. Running the PyR@TE-confirmed two-loop SM RGEs from M_Z up with the measured (m_H, m_t) gives λ(M̄_Pl) ≈ 0.002 with β_λ ≈ 0 — the celebrated Standard-Model near-criticality, now explained as a consequence of the free seam. The double condition predicts m_H ≈ 129–134 GeV (measured 125.25 sits a few GeV below, the known slight metastability); the same condition at the scalaron scale gives ≈107 GeV (too low), so the boundary condition lives at the Planck scale — consistent with seam = horizon = Planck (v166).

The source-level Koide quotient from the lepton φ₀-ladder is 0.664, 0.33% below the democratic compiler target 2/3 = |ℤ₂|/N_fam. A source→pole transfer conjecture brings it onto 2/3, but is not a derivation. The relaxation now has a canonical generator — dq/dt = (Δ/N_fam)·det B(q), the gap times the anchor-block quadric, whose time-1 map is the forced Möbius attractor — and the discrete-vs-continuous question is experimental: n = 3 = N_fam transfer steps corresponds to m_τ = 1776.9427 MeV (+0.14σ; n = 2 excluded at −2.9σ), decidable at σ(m_τ) ~ 0.01 MeV.

The candidate is the determinant-line axion of the strong-CP sector; WIMPs are ruled out (no spare E₈ singlet). The misalignment angle is closed; the decay constant is a conjecture. A misalignment estimate (v185) and a converged FULL finite-T solve (experiments/ftransfer/axion_relic/full_finiteT_solve.py: exact nonlinear misalignment, lattice χ(T)∝T⁻⁸·¹⁶, realistic g_*(T), normalised so θ_i=1 gives the standard Ω_a h² ≈ 0.03) now decide the abundance: at the predicted θ_i ≈ 170° hilltop the relic is Ω_a h² ≈ 0.66 — ~5.5× above Ω_DM h² = 0.12 (the observed value is reached only at θ_i ≈ 106°). So as the dominant dark matter the determinant-line axion at (f_a = M_scal/128 ≈ 2.39×10¹¹ GeV, θ_i ≈ 170°) OVER-closes the universe unless there is extra dilution or a lower f_a — a confirmed tension, not the optimistic all-DM. A more robust angle is the spine branch θ_i = π·N_fam/g_car = 3π/5 = 108° (v211): the same solver reaches Ω_DM at θ ≈ 106°, and 108° (the central spine quotient 3/5, no fit) sits there in the MILD-anharmonic regime — 62° below the hilltop, so NOT exponentially sensitive. It is an alternative ansatz to θ_i = π(1−φ_seam) ≈ 170° (mutually exclusive, the full solver decides, DM.AXION.SPINE.01) — a sharper [C] scenario, not a derivation; a converged Ω_a h² outside ~[0.08, 0.16] demotes the branch. That spine angle is exactly the regular pentagon interior angle: since N_fam = g_car − 2, θ_i = (g_car−2)π/g_car, so cos θ_i = (1−√5)/4 = −1/(2φ), and the golden character is unique to g_car = 5 (v429) — the otherwise-unmapped golden/icosahedral E₈ signature (v354/v313) is the geometry of this one external input, a [C] bridge that does not upgrade DM.AXION.SPINE.01. The haloscope coupling is tied to c₃: in the determinant-line normalization the axion–photon anomaly coefficient is g_aγγ = −4c₃ = −1/(2π), y² = 16c₃² = 1/(4π²) ≈ 0.0253 — the same c₃ that fixes α and the birefringence, with no flow freedom (v207); a [C] structural relation (the coefficient, not a parameter-free g_aγγ in GeV⁻¹, which still carries f_a).