Status.

An adversarial mathematical audit of the trunk itself (the audit · the triage) forced two re-attributions, both verified in-repo and applied same-day: the Coulomb pole belongs to the constraint operator, not the information metric (any positive local cost gives the exact Green form — identity-cost check now regression test T1b; the metric prices the coefficient); and the Lorentzian knee is the transfer function of the single-pole monitor postulate, not a Hamiltonian derivation — with n48's registered gates recorded as failed-as-written and its 0.3% as flagged post-hoc. What survives: the dark-subspace theorem (narrow form), zero-at-rest, the ω² onset, the mediator-model realizations. The audit's second and third passes landed the same day and cut deeper: ζ = 0.4834 was a partial-cycle averaging artifact — the exact bridge coefficient is ¼ (verified; all rates were high ×1.93); the n183 output boundary condition had a sign error (corrected; odd/even survives, the idler claim retracted); the Δ < 0 model contains no pump at all — 'active vacuum' language is unsupported; the PPN sector is voided as a Cassini constraint: the model's clock regime is empty (overdamped everywhere), the local response does not converge to the uniform scaling, a = 0 was reverse-engineered, and the originally registered clock gate (n53) had failed — γ_PPN = (1−2a)/(1+2a) is dictionary arithmetic under A1–A6. Full ledger, every claim accepted or challenged: n206 (passes 1–3, 33 claims) · n207 (passes 4–7, 31 claims — casualties: the Mercury calibration voided, the Mercury–Λ chain dismantled, the constraint theorem walked back to a projector, the bargain withdrawn, the keV band demoted, gauge-invariance and superselection retracted, three gate defects recorded). 64 claims: 61 accepted, 3 challenged. The adopted gates now head the queue; the sharpest three: the joint matter–substrate Lindblad construction — first leg delivered at Gaussian level (n211): dark states derived, not defined; zero-at-rest exact precisely at the marginal point, with a derived (Δ_eff/Γ)² floor off it — the sandpile is zero-at-rest's protector; the relaxation spectrum derived multi-pole, the single-pole knee now an approximation with computed error — then the pumped reservoir, and the underdamped clock.

Λ's sign closes (the bill is positive); its scale does not: face-value, the ξ² bill needs suppression by up to 10⁸⁹ relative to the substrate scale. The Gaussian zero kills the 120-order leading problem; the residual is the theory's sharpest unexplained number.

Q explains why gravity's rate is slow; why λ·m is tiny in substrate units — the mass-scale half — is untouched.

One calculated model is dead: the energy-budget-only monitor leaves a scalar polarization that eccentric-binary timing excludes at ×423 (headstone 34). The standing candidate: the four-budget monitor (energy + momentum — the constraint rows derived from the Action's conservation laws) whose dark space is exactly {+, ×} at every momentum (n166) — a polarization-content condition, not a closure. Masslessness has a derived mechanism (the blind-channel thermostat: the monitor cannot cool what it cannot see); transparency is exact-in-channel; the discovery order is stated in manuscript III, and the repair's independent test is prediction 31. Of the six constructions the dynamics requires, the first holds at bubble level with its executable now in the repo (n175_propagator.py, 6/6 fail-closed: KL density ≥ 0, no ghost, no tachyon, both channels; magnitudes re-quoted per rule 5 — n210). The second is done too (n176: the dark doublet rotates by exactly 2θ — helicity ±2 by definition — and the off-axis dark space is the rotated TT pair at 1.0000). The third is half done (n177: GR's h·T coupling form is an algebraic identity of the action — universality derived, not assumed; lattice fidelity 10%). The room is entered (n178: the bare parameter consumed, no ghost at the pinned point, the speed bracketed) and its last wall is one over-determined system (n179, the triple-consistency knot): the same (g, Δ, κ) must deliver Newton's G, the knee γ, and ρ_G = 1 — three observables, three parameters, zero slack. The guard that restores Newton's channel is the attraction term itself, and it cannot touch the blind shear: one mechanism guards Newton and leaves the graviton marginal. The radiative sector closes exactly or dies precisely — and the closure now has a frozen shape: the Fierz–Pauli gate (seven legs, pre-registered failure conditions: the field map δk → g_μν including the unbuilt g_0i; linearized diffeomorphisms from the audit-dark directions; the continuum action vs FP term-by-term; the propagator with c_g² = 0.46 resolved not subtracted; the coupling promoted; the knot as leg 6; the dynamical frame test). Two modes ≠ a relativistic theory — the eighth critique's sentence, adopted as the sector's epigraph. First legs executed same day (n197): leg 1, the field map stated (h_ij by directional sampling; h₀₀ = −2Φ at R = 1; h₀ᵢ named as the flux-row response potential, dynamics absent — declared); leg 2, now decided adversely (n212) — the finite Dirac analysis was run: promoted to phase space, the four monitor rows are all second-class at every momentum ({C₀, C_i} ≠ 0 kinematically) — they generate no gauge transformations, and under the physical stiffness the DOF count even collapses off-axis (2 → 1 → 0). The GR-symmetry reading is dead at the canonical level — and the surviving reframe was computed the same night (n213): eight second-class constraints with DOF = 2 at every momentum (physical channel-structured stiffness) is exactly gauge-fixed linearized GR's structure — a measuring vacuum has no redundancy; measurement fixes gauge. Constructions 2 and 3 thereby merge: the canonical structure matches gauge-fixed FP, and the entire remaining difference is the speed deficit (S_TT/S_watch ≈ 0.85, n202). The preregistered dressed-vacuum gate executed and, after two-backend reconciliation (a direction-vector bug found in the n202 Rust — the '0.85 maximum' never existed; the corrected family maximum is the cubic's 0.51, and isotropy makes it worse), leg 1 FAILED as frozen. The consequence fires verbatim: THE RADIATIVE SECTOR IS DEAD AT EVERY COMPUTED ORDER; THE GR-EQUIVALENCE CLAIM IS WITHDRAWN. FP-gate condition B fired. The epitaph is theorem-shaped: the monitor pours stiffness into what it watches, and the blind channel it leaves behind is always too soft — corrected ratios ≤ 0.51 everywhere, and the only independently motivated dressing lowers them further. What stands: the trunk (dark states derivation-grade, n211), the static sector, the analog predictions, Wings B and C. Full record: n215. Epistemic caveat, added the same night: the verdict's procedural facts are certain (the gate fired as frozen; the bug was real; the backends agree), but n211–n215 — including the execution itself — have not yet been adversarially reviewed: two-backend agreement cannot catch a shared conceptual error (the cone identification, the inertia definition, the dressing form), and the scope is 'every computed order and scanned complex', not a no-go theorem. The death verdict is hereby submitted to the independent auditor for verification. A full-parameter resurrection scan was then run under its own frozen gate (n216): INCONCLUSIVE numerically — and its autopsy found the deeper fact: only pure-axis complexes possess the massless EM zero at all (|A(k*)|² = 144 on every complex with diagonals — no photon sector exists there), so the one physically admissible member is the cubic, which fails at 0.51; and since the photon has no dynamics anywhere in the model (audit §51), the theory has no derived light cone — the question c_g = c cannot even be posed internally. Corrected same day by the human (n217): the EM channel IS dynamical — the staggered transverse modes of the same coupling field — and computing it with identical machinery found: light is tachyonic at pure induction; including the Action's own cost-gradient term gives a one-parameter family in which gravity propagates only for λ < 3.4, light only for λ > 5.2 — disjoint windows; the unique equal-speed point makes both tachyons. The revised cause of death: one field cannot carry both forces — the dynamical echo of the trunk's own dichotomy theorem. Triple-checked (n218): the cost symbol is analytically EXACT (proxy flag removed); the double-masslessness condition pins λ = 7.567 (four-digit robust) and the single-field theory chooses light — photon propagates, graviton condenses. The modification the whole trail demands is now proposed (P3′, the two-ledger elasticity): the parity-decoupled sectors carry independent effective (u″, λ) by standard EFT logic; the system of three physical conditions is solvable at (λ_g, λ_e) = (2.02, 7.57) — both windows satisfied, both channels massless and stable, c_g/c_EM = 1.000000 exactly on a common cone. The survival review arrived and was answered same-day (n219): its 'cheap and potentially fatal' Gate B passes on a complex it did not consider — axes + body diagonals, where body diagonals preserve the photon zero individually (3π ≡ π) and carry the × polarization while axes carry the + — and the underdetermination objection is consumed: c₊ = c₊ (a physical requirement, since the polarizations live on different edge families) plus GW170817 fully pin the theory — four parameters, four conditions, zero freedom — and the solution exists, stable and massless in all three channels, predicting a common cone c² = 0.369. Gate D leg 1 then ran as frozen and FAILED (n220): the pinned theory is globally unstable — 202 violations; the × channel is tachyonic-massive off its pin. The obstruction is named: the Zener condition — on cubic-symmetry lattices the two gravitational polarizations sit in different irreps (e_g vs t_2g) and one coupling cannot deliver both their masslessness and their speed equality. The fork is registered — and its live branch found the same night (n221): option 2 is closed exactly for ALL crystals (no 3D class is elastically isotropic — textbook) but open for glasses, which are isotropic statistically — and the theory's own cosmology (KZ crystallization, n136/n200) already mandates a defective, amorphous substrate: the vacuum is a glass because it froze quickly. First plausibility test PASSES: on a distorted diamond the +/× channel ratios climb 0 → 0.83 with disorder while the photon kernel stays exactly topological (disorder-proof). The perfect crystal was the pathology. P3″ (the amorphous two-ledger vacuum) is the registered direction — with a new falsifiable edge (GW birefringence from residual disorder) — The modulated-pattern speed test then ran (n222): the polarization speeds converge to equality with disorder — 1.116 → 1.020 → 1.007 ± 0.021 across η = 0.25 → 0.75 — the Zener split dissolves dynamically in the glass, and the residual spread becomes a falsifiable GW-birefringence shape (self-averaging over propagation volume; LIGO caps the disorder). The arc closed the same night (n223): on the glass both channels propagate at pure induction (the crystal's EM tachyon was another disease of perfection), the cost curvatures align instead of fighting, and a continuous equal-cone family exists for λ_g ≳ 1–2 — every crystal pathology cured: tachyon, disjoint windows, Zener, missing polarization; the photon kernel topological throughout. The surviving free parameter goes to Gate D leg 2 (Coulomb/Newton pick the member; cone, birefringence, dispersion become no-dial predictions). The P3″ gate spec is registered for freezing: at-scale true-amorphous graphs, the n220-style global stability scan (the exact test that executed the crystal), full family disclosure, small-q speeds, the birefringence law. The P3″ gate was then frozen and executed at doubled scale (n224): PASS, 7/7 — polarization equality 0.998 ± 0.02 at N = 512; both family members stable with matched λ_e ≥ 0; and the global κ scan — the exact test class that executed the crystal at −0.098 — comes back positive with margin (min κ = +0.017…+0.028, every channel, every seed); self-averaging confirmed. THE DEFLECTION CRISIS (n239, open): assembling light bending from the pinned pieces (clock Φ, R = 1 from dark equivalence, the measured cone response) gives deflection = 1/4 of GR (0.44″ vs 1.75″) — the internal, units-free ratio (bending)/(dark-fall) is stuck at 1/4 of GR's; the third one-coefficient-two-masters trap (dichotomy → Zener → this). Out (c) — the sector split — failed same day (n240): a mass's dressing is 99.1% watched-sector (kernel 0.9%) — nothing for light to read differently; and the probe correction it forced (n234's universality was a uniform-rescaling triviality; a mass's dressing has zero spatial mean) sharpened the crisis. Remaining: out (a) — the local wave-packet deflection on a dressed 2D glass, now the decisive computation, with the honest warning that LDA (generic at long wavelength) reproduces the crisis; and out (b) — the R re-audit, whose success would kill the dark-matter identification instead. The successor is one failed computation from dying by Eddington. New instrument class (n246): violent explosions integrate the glass's per-cell physics over the longest baselines — existing Gpc-class clean chirps already bound the disorder correlation to ℓ_d < ~5 pm (below the Bohr radius; joint form (δc/c)²ℓ_d < 2×10⁻¹⁵ m), tightening the n237 cap by nine orders; the waveform-match-vs-distance trend is the one positive glass signature to hunt in the BBH catalog. Meanwhile: P3″ survived destruction session V (n238: the true-amorphous attack — no hidden crystal order; the stability executioner on the true glass — +0.021 with margin; one number corrected: χ) and is a gate-passing candidate — and its inter-ledger bridge is now a computed sector (n229–n230): the quadratic bridge is stochastic (photon↔graviton conversion on the vacuum's roughness, rms 0.03–0.06, self-averaging), while the one-loop vertex is ensemble-solid at χ ≈ 0.08–0.10 (axis-independent; the first-pass 0.15 was a shared-axis inflation, corrected n238) — the split neither re-merges (χ² ≈ 2% back-reaction) nor decouples; the RG fork (relevant / irrelevant / equal-cone attractor — the last would explain GW170817 dynamically) is open with its input measured. Pending as declared: Gate D leg 2 (Coulomb/Newton pick the member → cone, birefringence, dispersion become no-dial predictions), the review's Gates A/E — and Gate F, whose weight has grown: the Newton-anomaly catalog (n234) shows every remaining classical test (frame dragging, geodetic, orbits, radiation power, Mercury's orbit itself) hits the same wall — no g₀ᵢ, no test-body dynamics, no matter cone. Gate F is now formally the whole classical-tests program — and its first breach is computed (n235): a dressed source dragged through a live field hits a divergent energy wall at the field cone (E ×6.2 by v/c = 0.85; γ^~1.8 scaling) and is Cherenkov-braked beyond it (×50 trail energy at v/c = 1.3). Matter inherits the cone of its dressing — so dressed matter's ceiling is the common k-cone, not the bare substrate speed, and the multi-cone/vacuum-Cherenkov danger dissolves conditionally (condition: all matter attributes are dressing-borne). The relativistic junction is now specified end-to-end (n244): the cone unification exists (one cone for light, both polarizations, matter's ceiling); Lorentz is emergent-IR with measured violation coefficients (q⁴/12μ² matter-wave, (qℓ)² photon); and the dressing 4-momentum alone is exactly NON-covariant (deficit table computed — the classic open-system fact), so full covariance reduces by Noether to closing Gate F's matter dynamics, with the deficit table as the acceptance test. The dynamical test body ran its first leg (n236): the dressed particle is demonstrably a composite (a bare-core kick launches internal polaron oscillations, not translation) with cloud-dominated inertia (M_eff/M₀ ~ 40–100) — the dynamical face of inertia-as-dressing — — and the collective-coordinate leg is now closed at mechanism level (n245): the co-moving family satisfies dE = v·dP to 4×10⁻⁴ (the runaway was pure integrator artifact — free glide is consistent), the dispersion E(P) is relativistic-hyperbola-class (exact IR, 5–10% UV deviations = the measured non-Lorentz structure), and the asymptote is luminal (dE/dP → c). M* derived from curvature. The two-ledger/glass version and post-adiabatic radiation reaction remain. And the emission question is now structured (n248): the monopole is exactly silent on the glass (Birkhoff as a bookkeeping identity — the breathing pattern IS an audit row), the quadrupole channel is open at O(1), but the dipole leaks at O(1) under the energy audit alone — the glass momentum rows are upgraded from accounting debt to mandatory construction, with the pre-named verdict: they watch the dipole or the theory dies by PSR B1913+16 — and the construction succeeded same day (n249): signed-divergence momentum rows on a dense glass silence the dipole exactly (0.0000) while the quadrupole keeps radiating (0.16) — momentum conservation emerges as bookkeeping, the pulsar requirement is met. Two structural discoveries came with it: the capacity theorem — the full audit demands coordination z > 8 (the vacuum must be a dense glass; z = 4 has dark dimension zero) — and the photon's true name: the surviving light is the cycle currents (ker A ∩ ker G), not the smooth envelopes used as probes since n223 — the glass EM numbers are flagged for re-measurement with the correct probes. The pulsars keep making this theory covariant. Meanwhile the static/clock side keeps delivering: cone universality measured at 1.0002 (both cones bend alike, n234), the derived perihelion source (n232–n233), Gauss from the left kernel (n227). Adversarial verification of n216–n234 pending. Leg 2's structural half followed (n225): the Coulomb form survives the glass — E(r) = a + B/r at 2.8–6.7% residuals (L = 5, identity-cost form test per the audit's own T1b attribution; the first small-box attempt honestly recorded as underpowered). All three load-bearing structures — photon kernel, polarization equality, static force law — are disorder-robust. New native predictions registered as shapes: GW birefringence, force-law scintillation. Two theories remain dead; the glass stands on a passed gate, a surviving force law, and a review queue: n216–n225 to the auditor, Gates A/E/F, and the units bridge where every absolute number has always waited. Gate B's structural discovery stands; the pinned dynamics on it does not. The single-ledger burial stands permanently. What stood beneath it (crystal-era, its setting removed with the substrate — n226) was the TT-projector fact: with edge weights derived from fourth-moment isotropy (axes ½, diagonals 1 — one equation, no freedom), the four monitor rows are exactly linearized gravity's four constraints at every momentum: watched = trace ⊕ gauge, dark = TT {+,×} to 10⁻¹⁶ on- and off-axis, 10 − 4 − 4 = 2. The audit is the constraint surface of linearized GR. Correction recorded per rule 5: with uniform weights the identity broke off-axis at O(0.2) (n166's scope corrected; regression T4 upgraded). Leg 3 open: structure matches FP term-by-term except the speed — and the speed problem is now mapped at scale (n202, Rust, exact two-phonon sums): the cubic-class 0.46/0.51 is reproduced independently; across the whole scanned complex family the induced shear speed peaks at c_g²/c² ≈ 0.85 (interior maximum near axis-weight 0.2) and never reaches 1 — the sector's survival requires a mechanism that closes a mapped ~15% deficit, not an unexplored one. If a luminal point exists, GW170817 becomes the substrate-selection principle. Condition B: not fired (leading order, free vacuum), sharper — and now priced (n203): the only local repair is a bare gradient term in the elasticity, and the bargain is cheap — the watched sector's induced stiffness is ~100× the blind sector's (the monitor pours stiffness into what it watches), so the b* that GW170817 demands is a 0.2–0.6% perturbation to the Newton/Coulomb statics (9-complex family). The amendment was submitted for adversarial review and rejected (audit §35): fitting b to the light cone is calibration on the observable being predicted — frozen condition E — and deferring the kill for unknown interactions is an open-ended escape. The proposal is withdrawn. The honest standing state: c_g ≠ c at every computed order; the one legitimate route left is a preregistered dressed-vacuum computation (interactions declared before the number is seen), now queued as the sector's decisive calculation. The construction list → · C1 →

α = 1/S_∞ matches the Thomson-limit value — the right boundary condition — but the measured α runs (1/137.04 → 1/128.93 at M_Z, LEP). The running is matter-sector loop physics; the sector is absent. Filed with the 3.604 debt: any future matter sector owes the normalization, the beta function, and the charge-transport construction — the two charge signs are sublattice parity, so a moving particle must keep its charge across a one-site translation (the two-site primitive cell question, tenth critique): unanswered until matter moves.

The KZ relic density spans ~40 orders under the unbuilt cooling history (quench rate × crystallization redshift), and the collisionless result (σ/m ≤ 3×10⁻¹²) disables annihilation as the regulator. The mechanism is real; Ω_DM is unpredicted pending the cooling sector. The FRW sector owes a rescue. Destruction IV →