Gate: Gap-13 — Black-hole entropy ($S = A/4G$) / Page curve, Quantum / gravity sector of the scoped theory on the frozen 13D $K_6$ branch. Frozen branch (READ-ONLY):
dcc66f1b2685/ manifest metaa5b1e6f9d951. Honest status (binding — matches the live popup chip): OPEN (DIAGNOSTIC) — one banked area-law value-match (the Bekenstein–Hawking value $S = A/4G$ recovered to 0.0028% relative error as a labeled consistency check) plus a fully-mapped downstream cascade. Honest, not closed. Direction: strengthened. Discipline carried verbatim: STATUS-UPGRADES:0. No coefficient $1/4$ asserted as a derived output. No quantitative Page curve claimed. given-E ≠ derivation of E; ANCHORED ≠ DERIVED; AXIOM-CLOSED ≠ atomic; a captured log ≠ an independent reproduction. The trap is anchoring on the target.
Headline. Black-hole entropy stays honestly OPEN — but it is not a dead end. The same frozen geometry already reproduces the Bekenstein–Hawking value $S = A/4G$ to 0.0028% relative error as a labeled consistency check, and the two things still missing — a microstate count and a Page-curve mechanism — are named, gated, and waiting on two specific upstream parents, not lost in a mystery. A focused completion campaign on this gate did something rarer and more valuable than a forced closure: it ran the gating theorem and got a refutation, which made the gate more honestly open by exposing exactly which one posit is the irreducible seam; and it then built a genuinely non-tautological conditional route to the coefficient $1/4$ (the conical-defect Einstein–Hilbert replica saddle) whose single remaining block is a precisely-named, literature-level heat-kernel coefficient shared with several other gates.
The honest grade. OPEN (DIAGNOSTIC). This dossier does not upgrade it. Under the program's grading rule (gate status = the least-closed residual; one open piece ⇒ gate OPEN), Gap-13 is OPEN because its two physics legs (R1 the static coefficient, R2 the Page mechanism) and the prerequisite they ride on (R6 the horizon-capable gravity completion) are all non-terminal OPEN. The ceiling is the program ceiling: serious candidate / partial unification — NOT validated.
What this dossier establishes, and what it does not.
The honest edge (a bet you can hold us to). This gate carries no falsifier of its own yet, by design: a five-clause anti-overclaim wall (PC-1…PC-5) forbids registering a Page-curve claim until a real mechanism exists, so the program cannot be caught faking a Page curve. The moment Gap-01 delivers the named near-horizon heat-kernel coefficient and Gap-14 delivers a consistent system–bath split, two sharp tests fire: (a) does a geometry-native state count give $\log N(A) = A/4G$ + known subleading with no fitted prefactor, and (b) does the radiation entropy $S_{\rm rad}(t)$ actually turn over at the Page time? Either a clean turnover or its refutation is a valid, publishable result. The one thing genuinely unprovable here is unprovable for everyone: that "no simpler horizon-capable formulation could ever exist" is a universal negative no theory in physics can discharge — a limit on all knowledge, not a hole in ours.
Every candidate quantum theory of gravity faces two demands at a black-hole horizon.
Static entropy — the coefficient $1/4$. Bekenstein (1973) and Hawking (1975) established that a black hole carries an entropy $$ S_{\rm BH} = \frac{k_B c^3}{\hbar G}\,\frac{A}{4} = \frac{A}{4 G} \quad(\text{Planck units}), $$ proportional to the horizon area $A$, not its volume, with the locked coefficient $1/4$. The demand is not to quote this number but to derive it: to exhibit the microscopic degrees of freedom whose count $N(A)$ satisfies $\log N(A) = A/4G + (\text{known subleading})$, "for the right reason," with no prefactor tuned to the answer. This is the statistical-mechanical foundation of horizon thermodynamics.
Dynamics — the Page curve. Hawking's semiclassical calculation gives radiation whose entanglement entropy rises monotonically to the full black-hole entropy and beyond, signalling information loss. Page (1993) argued that a unitary evaporation must instead produce radiation entropy that rises and then falls, peaking at the Page time (roughly when half the entropy has radiated) and returning to zero — the Page curve. Reproducing this turnover from first principles is the operational content of "information is not destroyed."
The honest community position: nobody has done either leg from first principles in a realistic 4D gravity theory. Gap-13 inherits that difficulty and does not pretend otherwise. Its contribution is to locate the difficulty with unusual precision and to bank one checkable diagnostic.
This section is the load-bearing technical core. It has four parts: (3.1) what the frozen geometry genuinely supplies on the gravity side; (3.2) the banked area-law value diagnostic and exactly what it is and is not; (3.3) the gating theorem T1 and its refutation (the static-leg adjudication); (3.4) the conical-defect non-tautology route T2 and the island/QES construction T3.
The frozen active branch is $$ \mathfrak{B}_{\rm active} = \big[M_4 \times K_6 \times S^2 \times S^1_Y/\mathbb{Z}_2\big]\ \oplus\ \big[F^+_{\rm finite} \oplus C_{\rm admiss}\big]\ \otimes\ \big[E_{\rm matter}\oplus E_{\rm gauge}\oplus E_{\rm Higgs}\oplus E_{\rm proton}\big], $$ with the internal coset $K_6 = SU(3)/T^2$ and the spin-$\mathbb{C}$ index $\chi(K_6, E) = -3$ (frozen). The 4D graviton arises as the normalized zero-mode reduction of the spacetime-facing block: $$ \text{frozen active branch}\ \xrightarrow{\text{zero-mode reduction}}\ \text{4D massless spin-2}\ \xrightarrow{\text{linearized}}\ \text{vacuum} + \text{masslessness} + \text{Newtonian limit} + \text{Einstein–Hilbert IR}. $$
This is genuine, geometry-essential machinery. It is the carrier of gravity: it delivers the linearized, weak-field, infinite-area sector — vacuum, masslessness, the Newtonian limit, and the Einstein–Hilbert IR as a route statement. Its measured floor is E (the linearized-GR / Einstein–Hilbert limit, carried as given-E input) plus the measured Newton constant $G$ supplying the zero-mode normalization.
Where it stops — the honest boundary. Horizon physics is a nonperturbative, strong-field, finite-area regime: $$ S = A/4G \text{ needs: a microstate count OR a horizon entanglement entropy}; \quad \text{Page curve needs: an island/QES OR a replica-wormhole saddle}; $$ both presuppose a horizon-capable nonperturbative / holographic completion of the gravity sector. The zero-mode projection does not reach this regime, and there is no automatic continuation in the frozen corpus from the perturbative carrier to the finite-area horizon. The carrier the horizon question is asked on is therefore an inherited route statement — and the graviton gate (Gap-01 / the GRAVITON completion) is itself CERTIFICATE-CONDITIONAL / OPEN. This caveat is carried throughout: the diagnostic below rides on a sector that inherits the linearized-GR limit as given-E input.
The conflation firewall (must be held). The framework contains an area law — the $SU(3)_c$ Wilson-loop confinement area law (AUDIT-tier, lattice-verified, not proven). This is QCD confinement, a static potential between color charges — not horizon entropy on a causal horizon. The two are physically unrelated; the corpus does not link them; borrowing the coefficient $1/4$ from anywhere in the confinement sector would be illegitimate and target-fitted. This firewall is residual R4, terminal.
On the 4D zero-mode sector of the frozen 13D geometry, the theory reproduces the Bekenstein–Hawking value $S = A/4G$ to 0.0028% relative error under freeze-before-compare discipline. This is banked as S1, graded DIAGNOSTIC: an inherited consistency check, not a derivation ab initio.
Precisely what S1 is: - It is a value-match: the numerical $S = A/4G$ computed in the inherited 4D effective-gravity description agrees with the Bekenstein–Hawking target to $2.8\times10^{-5}$ relative. - It is logged as an inheritance check: because the 4D sector inherits the Einstein–Hilbert IR (given-E), the area-law value follows by the standard semiclassical relation. The agreement confirms the inheritance is internally consistent; it does not count the microstates whose log is $A/4G$, and it does not supply the coefficient $1/4$ "for the right reason."
What S1 is not (the bright-line, carried from the frozen docs): - It is not the microstate-counting theorem (S2.c / MO-13-1). That theorem is explicitly OPEN. - It is not evidence that $1/4$ is a derived output. Per the analysis in §3.3, the entanglement route returns $1/4$ tautologically; the value-match shares that caveat. - Its own reproducibility witness is itself an open audit item (S1.b / MO-13-7): the derivation that yields the 0.0028% figure (effective action, graviton/zero-mode reduction, $G$ normalization, comparison target) currently lives in the external Gap-13 dossier and is not reproduced in the rendered corpus. Marked OPEN as an audit item. Its absence weakens only the diagnostic's auditability, not the gate's openness. We do not reconstruct or re-state the 0.0028% number's internal derivation here, because it is not in the rendered corpus to synthesize — see §6, Hole H5.
The honest framing on the public capsule line: "$S_{BH} = A/(4G)$ is inherited on the 4D zero-mode sector at 0.0028% relative error — DIAGNOSTIC, labeled; inherited, not derived ab initio." (Source: TOE_FINAL_merged.html, Gap-13 capsule; COST_FLOOR_WALL_IMPACT_LEDGER.md rows for TOE-GAP-13.)
The static leg asks: can the coefficient $1/4$ be derived (not quoted)? The campaign's strongest contribution was to run the proof obligation that would have answered "yes" — and it returned no, on rigorous grounds. Naming it a theorem is the point: running the gating theorem made the gate more honestly OPEN.
The claim T1 tested. "For the finite, species-fixed 4D zero-mode field content of the active branch, regulated by the inherited cost-floor resolution $\ell_*$, the species-summed vacuum entanglement entropy across a local causal cut equals the full Wald/Bekenstein horizon entropy scheme-independently — so the locked value $1/4$ is an OUTPUT of the spectrum, not merely guaranteed by construction."
Verdict: the strong (scheme-independent) form is REFUTED, on two independent rigorous grounds.
What survives — and why it does not close the gate. A strictly weaker shared-counterterm identity (T1′) is proved but is scheme-DEPENDENT: the leading area entanglement term is rendered finite by precisely the gravitational counterterms, so the leading term is the renormalized Bekenstein–Hawking term (Susskind–Uglum). (Cooperman–Luty, arXiv:1302.1878.) The algebraic consequence is exact and damning for any "derivation" claim: because the same heat-kernel coefficient renormalizes both $1/G$ and the entanglement cutoff term, $$ \frac{(\text{entanglement-area term})}{(\text{induced } 1/G)} = \frac{1}{4}\quad\text{for any minimal content — including a free scalar with no microstates.} $$ This is the genuine $\kappa^3/\pi$ true-by-construction signature: a target-blind computation returns the target tautologically, carrying zero independent evidential weight. (A non-minimally-coupled scalar leaves an unabsorbed, sign-indefinite area term with no clean statistical meaning — Kabat — so even the tautology is scheme-sensitive across the species sum absent an unproven cross-spin cancellation; that cancellation is named, not asserted.)
This is reduce, not relocate: T1 lowers the count of promotable claims (it disproves the one that would have promoted the static leg) and pins the irreducible seam rather than shuffling the mystery into a new premise. The seam it pins is the posit AX-ENT-EXHAUSTS ("horizon entropy is exhausted by vacuum entanglement entropy across the cut"): T1 proves this cannot be demoted to a theorem, and that any value-free posit engineered to output $1/4$ through it is target-fitted.
Verification provenance (carried for honesty): each load-bearing literature step was checked against the named primary sources; the decisive refutation step (gauge-EE ambiguity is an area-coefficient ambiguity) is cited explicitly to Casini–Huerta / CHMP. Two non-load-bearing imprecisions are disclosed: (a) arXiv:2204.09464 is authored by Huerta–van der Velde (audit-flagged correction applied); (b) the exact $(d-2)\xi$ prefactor and the linearized-graviton sphere-EE coefficient were not independently re-derived — the verdict needs only the existence of an unabsorbed term, not its value.
T1 closes off the entanglement route to $1/4$ as tautological. The campaign then built the route that is not tautological: the gravitational-action (conical-defect replica) route, plus the corresponding Page-curve island construction.
The route is: $$ \text{frozen 13D zero-mode gravity}\ \Rightarrow\ \text{4D horizon-capable Einstein saddle}\ \Rightarrow\ \text{replica / conical-defect entropy}\ \Rightarrow\ S = A/4G, $$ and it bypasses the shared Susskind–Uglum matter counterterm entirely. That is the point: the coefficient could have come out wrong if the saddle, conical defect, boundary term, or effective-action normalization were wrong — so it is a real candidate, not a tautology.
The algebraic core (rigorous; independently re-verified in sympy, both conventions, 2026-06-25). Using $\ln Z = -I$, $E = \partial_\beta I$, $S = \beta E + \ln Z = (\beta\partial_\beta - 1)I$:
Convention A (on-shell Euclidean Schwarzschild + Gibbons–Hawking–York). $$ I(\beta) = \frac{\beta^2}{16\pi G},\quad E = \partial_\beta I = \frac{\beta}{8\pi G},\quad S = (\beta\partial_\beta - 1)I = \frac{\beta^2}{16\pi G}. $$ At $\beta = \beta_H = 8\pi G M$: $E = M$, $S = (8\pi G M)^2/(16\pi G) = 4\pi G M^2$. With $A_H = 4\pi(2GM)^2 = 16\pi G^2 M^2$, one gets $A_H/(4G) = 4\pi G M^2 = S$.
Convention B (conical defect / Fursaev–Solodukhin). The replica geometry $M_n$ carries a distributional tip curvature, $$ \int_{M_n}\!\sqrt{g}\,R = n\!\int_{M_1}\!\sqrt{g}\,R + 4\pi(1-n)\,A_H + O\!\big((1-n)^2\big), $$ where the $4\pi$ is 2D cone topology (tip integrated curvature $= 2\times$ deficit $= 2\cdot 2\pi(1-n)$): rigid, content-blind, theory-independent. Inserting into the action, $$ I_n = n I_1 - \frac{(1-n)A_H}{4G},\qquad S = (n\partial_n - 1)I_n\big|_{n=1} = \frac{A_H}{4G}. $$ Both conventions agree on the $n\to 1$ derivative. The decomposition is exact: $$ \frac14 = \underbrace{\frac{4\pi}{16\pi G}\cdot G}_{\text{(conical-tip solid angle)}\,\times\,\text{(EH normalization)}^{-1}\,\times\,(G\ \text{measured anchor})}. $$ No matter loop, no UV cutoff $\epsilon$, no entanglement-cutoff matching enters: $16\pi G$ is fixed by Newton, $4\pi$ by 2D cone topology. The coefficient is geometric / topological, not thermodynamic and not entanglement-derived.
The conditional theorem (PROVED). Under four value-free hypotheses,
$$ (H1 \wedge H2 \wedge H3 \wedge H4)\ \Rightarrow\ S = A_H/(4G), $$
with H1: controlled Euclidean continuation + a nonperturbative 4D induced-Einstein saddle continuing the perturbative carrier; H2: the replica family $M_n$ (Euclidean Schwarzschild/Kerr, conical defect at the bolt) admissible/dominant under the frozen $C_{\rm admiss}/F^+$ grammar, with the conical identity on the actual $S^1_Y/\mathbb{Z}_2$ orbifold boundary; H3: internal modes spectator and stable on $M_n$ (no defect-sourced internal negative/zero mode, no internal order-$A$ term, no shift of $G$); H4: induced effective action exactly Einstein–Hilbert at order $A$ with frozen $G$ (no unfrozen higher-curvature/Wald or boundary-defect order-$A$ term). None of H1–H4 contains $1/4$ or is tuned to $A/4$. The conditional implication is rigorous; the HORIZON-ADMISSIBILITY theorem — the discharge of H1–H4 — is NOT achieved. (Source: certificate G13_T2_horizon_EH_replica_saddle/CERTIFICATE.md; 09_HORIZON_ADMISSIBILITY_PASS_2026-06-25.md; T2_CONICAL_DEFECT_CERTIFICATE.md.)
The falsification test table (where $1/4$ could have failed). A route is non-tautological iff at least one load-bearing step could have failed and the number itself is not engineered:
| # | Load-bearing step | Could $1/4$ have come out wrong? | Status on the frozen branch |
|---|---|---|---|
| KT-1 | Saddle exists (HORIZON-ADMISSIBILITY) | YES — catastrophically (no saddle ⇒ no number at all) | UNPROVEN — whole route conditional on it (= R6). |
| KT-2 | EH normalization $G_{\rm eff} =$ measured Newton $G$ | YES (a $K_6\times S^2\times S^1$ volume/kinetic factor could give $G_{\rm eff}\neq G_{\rm Newton}$) | CONSISTENCY CONDITION, not independently verified for the exact reduction. |
| KT-3 | Conical coefficient $4\pi$ (Gauss–Bonnet tip) | NO (2D cone theorem, content-blind) | RIGID — cannot be tuned. |
| KT-4 | Replica/thermo operator $(n\partial_n - 1)$ | NO (definition of saddle entropy) | RIGID. |
| KT-5 | No order-$A$ higher-curvature/Wald term | YES (Wald correction shifts the leading coefficient) | UNVERIFIED — needs the actual 4D effective action incl. curvature-squared terms. |
| KT-6 | $S_{\rm BH} =$ saddle Gibbons–Hawking entropy (AX-SADDLE-ENTROPY) |
NUMBER: NO (no coefficient in the posit). ONTOLOGY: an assumption | CARRIED VALUE-FREE POSIT — clean replacement for AX-ENT-EXHAUSTS. |
KT-1, KT-2, KT-5 are genuine failure modes ⇒ $1/4$ could have come out wrong ⇒ non-tautological (the entanglement / Susskind–Uglum route could not fail — that was the $\kappa^3/\pi$ tautology). KT-3, KT-4 are rigid (zero evidential weight). KT-6 is value-free (no $1/4$ baked in). Net: $1/4$ is DERIVED-GIVEN-E conditional on KT-1 (unproven) + KT-6 (carried posit), with KT-2/KT-5 unverified ⇒ R1 stays OPEN.
The induced-gravity / "species" objection — answered. Under AX-INDUCED-G, $1/G$ is the matter vacuum response, so one might fear $A/4G$ hides the Susskind–Uglum counterterm inside $G$. Resolution: the route is non-tautological iff $G$ is supplied as the measured IR anchor (Case 1) — matter loops already resummed into the physical $G$, the entanglement identification never invoked; then the coincidence $A/4G^{(\rm SU)}_{\rm induced} = A/4G^{(\rm conical)}_{\rm phys}$ is induced-gravity consistency, not a conical-route tautology. (If instead one computes $1/G_{\rm induced}$ from the same loop — Case 2 — the objection bites and one is back on the dead route; the certificate explicitly takes Case 1.) This is exactly ANCHORED ≠ DERIVED: $1/4$ is anchor-transferred onto the measured $G$ + one value-free posit + one unproven theorem; floor $\geq 1$ preserved.
The sharp reduction (REDUCE, not RELOCATE). Hypotheses H3 and H4 collapse onto a single named object: the order-6 mixed Neumann/Dirichlet $S^1_Y/\mathbb{Z}_2$-orbifold-boundary + conical-defect Seeley–DeWitt coefficient. Integrating out the internal towers on $M_n$ is — by the warped-product structure — a sum over internal KK degeneracies of the 2D-cone-with-$\mathbb{Z}_2$-boundary heat kernel; the $K_6 = SU(3)/T^2$ spectrum and $\chi = -3$ fix the internal weights (frozen/derived), but the order-6 boundary coefficient they multiply is exactly the object absent from the literature (the boundary heat-kernel tower stops at $a_5$; Branson–Gilkey–Kirsten–Vassilevich). This is the corpus's shared boundary heat-kernel wall — Gap-01 (owner); consumers GRAVITON R1/R6, UQF-3 R4, UQF-9 R2, SG-6 R9 — count once. H1/H2 cascade onto UQF-3 R3 (Clay-class Euclidean existence / reflection-positivity) and UQF-9 (UV completion); the H3 internal-stability leg onto SG-6 (moduli stabilization).
Honest twin observation (carried verbatim). The feature that makes $1/4$ non-tautological ("could come out wrong") is identical to the feature that makes it unproven ("must compute the defect coefficient to know it didn't") — and that computation is the BLOCKED object.
Modesty flag (do not over-read the conditional). H4 ("induced action exactly Einstein–Hilbert at order $A$, frozen $G$, no Wald / higher-curvature / boundary-defect order-$A$ term") is near-coextensive with "the $A$-coefficient is $1/4G$." The theorem is saved from circularity only because H4 is recast as a value-free structural vanishing whose truth is fixed by the uncomputed, named coefficient. Most of the work is honestly relocated into H4 and then located (not eliminated). The content of the conditional is modest; this stays explicit.
Stale-prose correction (carried so the specialist does not repeat it). Earlier framing referenced an "uncomputed $a_6$ / ~31% $|{\rm Riem}|^2$ error" as if the bulk $a_6$ were still open. Per the current corpus the bulk $a_6$ is COMPLETE_CROSSCHECKED; only the DEFECT / boundary order-6 mixed N/D $S^1_Y/\mathbb{Z}_2$ coefficient remains BLOCKED. The load-bearing object is precisely that BLOCKED defect coefficient — say "defect/boundary coefficient," not "bulk $a_6$." (Source:
09_HORIZON_ADMISSIBILITY_PASS§5 flavor-text correction; cross-checked against the Gap-01 first-pass plug result below.)
For the dynamical leg, the campaign performed a spherical s-wave reduction of the frozen 4D Schwarzschild sector to a 2D dilaton-gravity throat, with the dilaton $\phi(r) = A(r)/4G$ as the area operator and matter $=$ $c$ free 2D fields (s-wave-reduced frozen spectrum) coupled to a bath. The generalized entropy $$ S_{\rm gen} = \frac{\phi(\partial I)}{4G} + S_{\rm bulk}(R\cup I) $$ was extremized symbolically and the Page comparison $S_{\rm no\text{-}island}(t) = S_{\rm island}$ set up.
The QES solves explicitly (a genuine, geometry-driven island shape). $$ x_\star = -\frac{r_h}{2} + \frac{\sqrt{3 G c + 9\pi\alpha\, r_h^2}}{6\sqrt{\pi\alpha}}; \qquad \text{semiclassical } (Gc \ll \alpha r_h^2):\ x_\star \sim \frac{G c}{12\pi\alpha\, r_h}. $$ The QES sits a distance $\sim Gc/r_h$ outside the horizon — the textbook island location, derived from the frozen-geometry inputs, not assumed. The plateau is area-dominated: $S_{\rm island}\sim \phi(r_h) = \pi\alpha r_h^2/G = S_{\rm BH}$ precisely when $\alpha = 1/4$. The Page transition has the standard form $t_{\rm Page}\sim 6\,S_{\rm BH}/(c\kappa)$.
Two validity conditions — one met in form, one fails (the decisive blocker).
- Condition 1 — area term from T2, not assumed: MET IN FORM, but inherits all of T2's conditionality. $\alpha = 1/4$ is kept symbolic and sourced from T2 (never assumed $A/4G$ — avoiding the "move the target into the QES formula" trap), so T3 adds zero independent evidential weight on $1/4$; it only shows the island machinery is consistent with the turnover given T2. The 2D throat is the s-wave reduction of the same 4D Euclidean horizon saddle T2 needs, so T3 is conditional on the unproven HORIZON-ADMISSIBILITY theorem (KT-1 / R6) + AX-SADDLE-ENTROPY. T3 cannot be more solid than T2.
- Condition 2 — central charge derived from the frozen spectrum: FAILS. The 2D central charge setting the Page-time slope is not a clean frozen invariant: (A) it is gauge-convention-dependent (2D gauge fields are not clean $c=1$ CFTs; $c\in\{26.5, 50.5\}$, a factor $\sim 2$ shifts $t_{\rm Page}$ by $\sim 2\times$); (B) greybody/angular-momentum-tower-dressed ($c_{\rm eff}$ is a greybody-weighted sum over all $\ell$, not the bare $\ell = 0$ count); (C) Hawking-temperature(=mass)-dependent (only $m\ll T_H$ fields radiate ⇒ $c\sim 2$ at solar mass to $\sim 50$ at Planckian, a factor $\sim 25$). There is no single target-blind frozen-derived $c$. So the Page time is not fixed by the frozen spectrum; only the qualitative turnover (rise → plateau at the area term) survives scheme-independently.
Net for T3. The Page leg (R2) is OPEN, now with an explicit QES/island construction (not merely qualitative-by-assumption): the island shape is DERIVED-GIVEN-E conditional on the T2 saddle; the Page time stays OPEN because a target-blind frozen central charge does not exist. Notably, no AX-HORIZON-FINITE-DIM is needed as an input now (the QES turnover is structural — a genuine improvement over the prior framing), but the magnitude is unfixed. It is correctly not phrased as a theorem of unitarity. (Source: 09_HORIZON_ADMISSIBILITY_PASS §9; analysis artifacts scratchpad/t3_qes_solve.py, t3_check_semiclassical.py, t3_central_charge_killtest.py.)
The progress on Gap-13 came from a small number of disciplined moves. They are now shareable, because they are the reasoning, not the engineering.
4.1 Run the gating theorem even when you expect it to refute. The single most valuable move was treating "entanglement entropy = Wald entropy, scheme-independently" as a proof obligation (T1) rather than a convenient assumption. The refutation is not a setback; it pins the floor. By disproving the one route that would have promoted $1/4$ to an independent output, T1 tells the specialist exactly which posit (AX-ENT-EXHAUSTS) is irreducible and exactly why a "value-free" axiom engineered to output $1/4$ would be circular. This is the program's discipline in action: a negative is a valid, load-bearing result.
4.2 Separate the number from the ontology. The entanglement route's tautology lived in conflating "what the entropy is" with "what its value is." The conical-defect route splits them. The number $1/4$ comes from rigid geometry (Gauss–Bonnet tip $4\pi$ over EH normalization $16\pi G$, with $G$ the measured anchor); the ontology (that $S_{\rm BH}$ is the saddle's Gibbons–Hawking entropy) is carried as a value-free posit AX-SADDLE-ENTROPY. This is strictly cleaner than the old AX-ENT-EXHAUSTS, which returned the number by construction. The lesson generalizes: when a coefficient keeps coming out "right," check whether the route could ever have produced it wrong; if not, it is true-by-construction and carries no weight.
4.3 The anti-target-fitting axiom. The campaign promoted the anti-tautology requirement itself to an axiom: AX-BLIND-CUT-MEASURE (the measure on horizon alternatives must be fixed by the frozen geometry / admissibility grammar before, and independently of, any comparison to $A/4G$). This is the $\kappa^3/\pi$ falsification test written into the floor. It is why the conical route is admissible (it never touches the shared counterterm) and the entanglement route is not.
4.4 Reduce by locating, not relabeling. The HORIZON-ADMISSIBILITY pass did not lower the wall; it found that two of its four hypotheses (H3, H4) are the same object — the order-6 mixed N/D orbifold-boundary + conical-defect Seeley–DeWitt coefficient — and that this object is already tracked as a shared wall consumed by Gap-01, GRAVITON, UQF-3, UQF-9, and SG-6. Counting it once across all consumers is a genuine consolidation: a single literature-level computation discharges the static leg of Gap-13 and contributes to four other gates. This is leverage, not bookkeeping.
4.5 The axiom floor is below entropy. Iterating "what are the simplest, most fundamental axioms?" to a fixed point, the irreducible primitive turned out to be not entropy, microstates, records, or entanglement, but boundary-local causal distinguishability (master axiom AX-BOUNDARY-LOCAL-CAUSAL-DISTINGUISHABILITY, with sub-axioms A1 causal-separability, A2 boundary-local, A3 finite-capacity, A4 blind-measure, A5 distinction-transfer). Crucially, A2 (the distinguishable alternatives live on the cut boundary) is what makes it an area law rather than merely a finite-entropy statement. This reframing dissolves the dangerous AX-ENT-EXHAUSTS seam — entropy is distinguishability, not assumed entanglement — and is why the static leg could be re-attacked without falling back into the tautology.
4.6 Build the island from the geometry, keep the area term symbolic. In T3, the QES position $x_\star$ came out at the textbook location $\sim Gc/r_h$ outside the horizon from the frozen s-wave-reduced spectrum, with the area-term coefficient $\alpha$ kept symbolic and sourced from T2. This is the discipline that prevents the Page-curve construction from secretly assuming $A/4G$: the island shape is a genuine geometry-driven output, while the one number it cannot honestly supply (the central charge $c$, hence the Page time) is flagged as the decisive open object.
The gate's openness is decomposed into a residual register and a 26-row missing-object inventory. Summary (sources: 01_DOSSIER.md §3; specialist_reply.txt §2; TOE_SELF_CLOSE_OPEN_ITEM_INVENTORY.md Gap-13 block):
| ID | Residual / object | Current honest status |
|---|---|---|
| R1 | No horizon microstate count / no $S=A/4$ derivation (static leg) | OPEN — STRENGTHENED by T1. Promoting strong theorem REFUTED; only scheme-dependent T1′ survives (κ³/π); $1/4$ now also DERIVED-GIVEN-E conditional via T2 (KT-1 unproven). NOT AXIOM-CLOSED. |
| R2 | No Page curve / no island-QES-replica machinery (dynamical leg) | OPEN. Explicit QES shape built (T3); qualitative turnover survives; Page time OPEN (no target-blind $c$); blocked on R6. |
| R3 | Routing gap — owned by no internal module | DISCLOSED / exported as a named owed-module obligation. Non-terminal. |
| R4 | Confinement-area-law conflation ($SU(3)_c$ Wilson loop ≠ horizon $S$) | TERMINAL — firewall. Kept unlinked. |
| R5 | Perturbative graviton scope boundary | DISCLOSED (scope) — route-given for the perturbative part; OPEN for the horizon. |
| R6 | Horizon-capable gravity completion is itself unbuilt | OPEN / global open problem — exported to the gravity-sector ledger (= KT-1 = HORIZON-ADMISSIBILITY). |
| R7 | Bekenstein/Hawking/Page are external TARGETS, not outputs | TERMINAL — given-E targets. Counting them as outputs is the signature mis-close. |
| T1 | Gating proof obligation (scheme-independent EE = Wald) | CLOSED — VERIFIED. Strong form refuted; T1′ proved but scheme-dependent. Terminal as adjudication. |
Missing-object highlights (from the 26-row Gap-13 inventory): MO-13-1 / S2.c BH microstate ensemble + counting theorem ($\log N = A/4G$) — OPEN, gated on Gap-01; MO-13-2 / S3 unitary-evaporation / Page mechanism — OPEN, gated on Gap-14 → Gap-01; MO-13-3 Gap-01's $\mathrm{tr}[a_6]$ — INHERITED, walled at field grade; MO-13-4 / S3.a Gap-14's system–bath split — INHERITED, CH-2 honest-halt (NotImplementedError); MO-13-5 / S2.b topological-rank → degeneracy map (PC-4) — RESERVED, deliberately inert; MO-13-6 discharge-gate registry — MISSING (governance honesty wall); MO-13-7 / S1.b reproducibility witness for 0.0028% — MISSING (audit-only); S1 area-law inheritance diagnostic — present (guarded), DIAGNOSTIC; S4 PC-1..PC-5 disclaimer ledger — closed-as-record.
A documentation/non-promotion certificate, verified internally consistent and to cover every open node and the diagnostic (truth-table item 27). Its collective force, assembled clause-wise (verbatim texts EXTERNAL in the dossier — see Hole H7):
(Source: TOE_FINAL_merged.html Gap-13 capsule; specialist_reply.txt §3 cross-walk.)
dcc66f1b2685 / manifest meta a5b1e6f9d951, READ-ONLY. The graviton arises as the normalized zero-mode reduction of the spacetime-facing block. No horizon-sector primitive, microstate count, island prescription, or replica saddle has any hash — none exists as a frozen object; the T2/T3 constructions are adjudication-overlay analyses, not frozen-corpus outputs.G13_T2_horizon_EH_replica_saddle/CERTIFICATE.md: certifies the conditional implication $(H1\wedge H2\wedge H3\wedge H4)\Rightarrow S = A_H/4G$ and the non-tautology of $1/4$. Verdict: VALID as scoped, NOT target-fitted. It certifies the conditional and the non-tautology, NOT a closed gate. Algebraic core independently re-verified in sympy (both conventions).T2_CONICAL_DEFECT_CERTIFICATE.md and 09_HORIZON_ADMISSIBILITY_PASS_2026-06-25.md: the full falsification test tables, the REDUCE-not-RELOCATE verdict, and the T3 QES/island analysis.| Axiom | Content | Atomic? | Disposition |
|---|---|---|---|
| AX-INDUCED-G | $1/G$ is the vacuum two-point response of the branch fields; no bare Einstein–Hilbert term | NO | Carried posit, floor $=1$. Proved not reducible to root GRANULARITY (a Regge/lattice theory can carry a bare EH term). |
| AX-ENT-EXHAUSTS | Horizon $S$ exhausted by vacuum entanglement entropy across the cut | NO | Carried posit; proved non-demotable by T1 — the weakest link. In the T2 coefficient argument it is retired and replaced by the value-free AX-SADDLE-ENTROPY. |
| AX-SADDLE-ENTROPY | $S_{\rm BH}$ is the Gibbons–Hawking entropy of the gravitational saddle | value-free | Clean replacement for AX-ENT-EXHAUSTS in the coefficient argument (fixes ontology, not value). |
| AX-U (unitarity) | Branch evolution exactly unitary | YES (generic QM floor) | Gives only the qualitative Page turnover; adds no gravitational ontology. |
| AX-HORIZON-FINITE-DIM | Horizon coarse-grains a finite-dim Hilbert space, $\dim = e^{S}$ | NO | Carried separately so its dependence is not hidden inside "unitarity"; presupposes R6, cascades on Gap-01. (T3's QES turnover no longer needs it as an input.) |
| AX-FLOOR-INHERITED (GRANULARITY / $\ell_*$) | Finite minimal cost-resolution $\ell_* \sim \Lambda_{\rm YM}^{-1}$ | n/a (inherited) | Bookkeeping; supplies the shared UV regulator. Does not yield $1/4$ independently. |
| AX-BOUNDARY-LOCAL-CAUSAL-DISTINGUISHABILITY | Master axiom: every admissible finite causal cut induces a finite, boundary-local, geometry-fixed, target-blind measure of distinguishable alternatives, preserved/transferred under admissible evolution | floor (converged) | The stabilized fixed-point floor (A1–A5); dissolves the AX-ENT-EXHAUSTS seam; A4 = the anti-κ³/π axiom. |
Net floor: two carried, non-atomic-but-irreducible-against-root posits (AX-INDUCED-G + AX-ENT-EXHAUSTS / its value-free successor AX-SADDLE-ENTROPY) over a generic unitarity floor, a separately-carried open-premise finite-dimensionality posit, and an inherited regulator (zero new cost). Floor $\geq 1$. No new tuning introduced or removed. The cost-floor / granularity axiom dissolves the continuum near-horizon UV-completion prerequisite (a fixed resolution $\ell_*$ declines the $a\to 0$ idealization at the horizon) — but this removes the prerequisite burden, not the count, and the printed framework verdict for the Gap-13 whole is untouched.
scratchpad/t3_qes_solve.py, t3_check_semiclassical.py, t3_central_charge_killtest.py).This is the most load-bearing section: for each open hole, a concrete work-package a specialist can act on immediately. Physics only; applications firewalled.
(a) Precise statement. Prove that the frozen 13D branch's 4D gravitational zero-mode sector extends from the perturbative graviton carrier to an admissible Euclidean black-hole replica saddle family $M_n$ (Euclidean Schwarzschild/Kerr with a conical defect at the bolt), admissible/dominant under the frozen $C_{\rm admiss}/F^+$ grammar, with internal $K_6\times S^2\times S^1_Y/\mathbb{Z}_2$ modes stable/spectator (except through frozen $G$) and no unfrozen horizon-local order-$A$ term. This is the discharge of hypotheses H1–H4 of the T2 conditional theorem.
(b) Why it's hard / prior-attempt lessons. It is the global open problem of building a horizon-capable nonperturbative gravity sector (R6), shared with all of quantum gravity. The campaign reduced it but did not solve it. Specific traps to avoid: - Do NOT phrase the Page curve as "a theorem of unitarity." It is a qualitative turnover; the magnitude is open (overstated-theorem guard). - Do NOT route any coefficient argument through the shared Susskind–Uglum $1/G$ counterterm — that returns $1/4$ tautologically (the entanglement route is dead, T1). - Stale-prose trap: do not say "uncomputed bulk $a_6$ / ~31% error" — the bulk $a_6$ is COMPLETE_CROSSCHECKED; only the defect/boundary coefficient is blocked. - Modesty trap: H4 is near-coextensive with "the $A$-coefficient is $1/4G$"; assert "no order-$A$ term" only as a value-free structural vanishing fixed by the computed coefficient — never assert it to make the theorem go through.
(c) Exactly what closes it. Target-blind and reproducibly: (i) establish a controlled Euclidean continuation and an admissible dominant nonperturbative 4D induced-Einstein saddle (H1/H2); (ii) certify internal stability on $M_n$ (no defect-sourced internal negative/zero mode); (iii) certify no unfrozen order-$A$ higher-curvature/Wald term (H4). Success criterion: H1–H4 all discharged ⇒ the §3.4.1 conical proof delivers a non-tautological $1/4$. A refuting result is equally valid: if no admissible saddle exists, the gate stays OPEN for a deeper reason (perturbative gravity but no certified horizon sector) — a publishable structural finding.
(d) Machinery & inputs. Euclidean quantum-gravity saddle-point methods; Fursaev–Solodukhin conical calculus; the frozen effective action from the 13D → 4D reduction. Start from 08_FOUND_AXIOMS_AND_BUILD_PLAN.md §3–§4, the certificate G13_T2_horizon_EH_replica_saddle/CERTIFICATE.md, and 09_HORIZON_ADMISSIBILITY_PASS_2026-06-25.md. Build the certificate artifact prescribed in 08 §7 (effective action; saddle-admissibility proof; internal-mode spectator/stability audit; conical-defect calculation; explicit no-Susskind–Uglum statement; falsification test table).
(e) Leverage. This is the shared boundary heat-kernel wall counted once: closing it (via the defect coefficient, H4 below) discharges the static leg of Gap-13 and feeds GRAVITON R1/R6, UQF-3 R4, UQF-9 R2, and SG-6 R9. Its H1/H2 legs cascade onto UQF-3 R3 (Clay-class) and UQF-9; its H3-stability leg onto SG-6.
(a) Precise statement. Compute the order-6 mixed Neumann/Dirichlet Seeley–DeWitt boundary coefficient for the $S^1_Y/\mathbb{Z}_2$ orbifold, evaluated on a conical-defect (replica) background, and show its order-$A$ contribution vanishes for the frozen internal content. This single object discharges both H4 (no unfrozen order-$A$ term / frozen $G$ on the defect) and the no-internal-$A$-term part of H3.
(b) Why it's hard / prior-attempt lessons. The object is absent from the literature: the boundary heat-kernel tower stops at $a_5$ (Branson–Gilkey–Kirsten–Vassilevich). It is finite-in-principle but unevaluated, and the geometry (cone tip $\times$ $\mathbb{Z}_2$ orbifold boundary, mixed boundary conditions) is genuinely novel. Trap: never fabricate or assert "no $A$-term" without computing the geometry-fixed value — that would make $1/4$ DERIVED-GIVEN-that-posit, not proved, and is refused.
(c) Exactly what closes it. Compute the coefficient with Gilkey/Vassilevich boundary heat-kernel machinery extended to order $a_6$ with mixed N/D conditions on the orbifold boundary and a conical-defect background; show order-$A$ vanishing for the frozen $K_6 = SU(3)/T^2$ + $\chi = -3$ internal weights. Success: vanishing confirmed ⇒ H4 (and H3-spectator) discharged. Refutation: a nonzero order-$A$ term ⇒ $S = A/4G + (\text{Wald correction})$, shifting the leading coefficient — a valid, decisive negative.
(d) Machinery & inputs. Gilkey's invariance-theory boundary heat-kernel expansion; Vassilevich's review machinery; mixed-boundary-condition Seeley–DeWitt technology; conical/orbifold heat-kernel methods (Fursaev–Solodukhin, Dowker). The internal weights are already frozen (do not recompute them). Coordinate with Gap-01 (owner of the shared wall). This is a perturbative-but-heavy computation, AI-setup-able but not AI-closeable.
(e) Leverage. The highest-leverage single computation for the static leg. Shared across Gap-01, GRAVITON, UQF-3, UQF-9, SG-6 — count once, discharge many.
(a) Precise statement. Prove that the count of horizon-localized states over the $M_{\rm GUT}$ degrees of freedom satisfies $\log N(A) = A/4G + (\text{known subleading})$, for the right reason, with no per-entry fitting. The 0.0028% value-match (S1) is explicitly NOT this theorem.
(b) Why it's hard / prior-attempt lessons. It is flagged plausibly external-depth (Strominger–Vafa-class) even granted the cascade. Without a UV-controlled near-horizon Hilbert space (Gap-01) there is nothing well-defined to count. The entanglement route to the count is dead (T1: it returns $1/4$ tautologically, so fabricating a count would be target-fitting — this is exactly what PC-2 forbids). Trap (verifier-caught in a sibling gate): the Gap-01 first-pass plug fabricated a graviton $\sigma$-supertrace weight "67" and a ghost weight "11" that appear nowhere in scripts or frozen corpus — the corpus consistently uses graviton dim 91, ghost dim 13, and the combination grav $- 2\times$ghost $= 65$, never 67/11. Do not import or invent such weights; do not target-fit; stay within what is computed.
(c) Exactly what closes it. After Gap-01 supplies a UV-controlled near-horizon Hilbert space, construct the near-horizon state ensemble from the frozen geometry, count it, and prove $\log N = A/4G$ freeze-before-compare clean. Success: $\log N(A) = A/4G$ + known subleading, no fitted prefactor. Refutation: a count that lands on a different coefficient, target-blind, is a valid negative.
(d) Machinery & inputs. The frozen spectrum (degeneracies of the $K_6/S^2/S^1$ towers at the horizon scale); a Gap-01-supplied near-horizon Hilbert space; state-counting / index methods. Start from the canonical inventory (TOE_SELF_CLOSE_OPEN_ITEM_INVENTORY.md Gap-13 block, S2/S2.a/S2.c) and the owner hints (07_OWNER_HINTS_AND_DIRECTIONS.md §A.1).
(e) Leverage. Closing it (with H1/H2) gives the first independent non-tautological handle on $1/4$ (the campaign's corollary T2 target). Consumes Gap-01; produces nothing Gap-01 needs.
(a) Precise statement. Compute the Hawking-radiation entanglement entropy $S_{\rm rad}(t)$ over the evaporation history on the 13D fields, showing it rises then falls, with the Page time set by the microstate count (S2 and S3 must agree at the turnover). The corpus currently makes NO statement that $S_{\rm rad}$ turns over in this geometry; only a qualitative, area-dominated plateau is established.
(b) Why it's hard / prior-attempt lessons. Triply gated: S3.a (Gap-14 system–bath split) + S2 (the count) + Gap-01. The explicit QES (T3) is built and the island sits at the textbook location, but the central charge $c$ has no target-blind frozen value (gauge-convention, greybody, and mass dependence give factors of $\sim 2$–$25$), so the Page time magnitude is unfixed. Trap: do not source the area term from an assumed $A/4G$ — keep $\alpha$ symbolic and source it from T2, or the construction merely moves the target into the QES formula.
(c) Exactly what closes it. After Gap-14 supplies a consistent system–bath split, compute $S_{\rm rad}(t)$ via the geometry-native island/QES saddle (T3, with the area term from T2) and show the decrease after the Page time — OR show it does not (a refutation is equally valid). Resolve the central charge target-blind (the decisive sub-object). Success: a clean turnover with Page time consistent with the count. Refutation: no turnover, or an inconsistent Page time — a valid, publishable negative.
(d) Machinery & inputs. The s-wave-reduced 2D dilaton-gravity throat ($\phi = A/4G$); the generalized-entropy / QES extremization; greybody-weighted central-charge accounting. Start from 09_HORIZON_ADMISSIBILITY_PASS §9 and the T3 scripts. Requires H1 (the saddle) and Gap-14's split first.
(e) Leverage. Closing it is the campaign's corollary T3; it is the dynamical half of the gate. Consumes Gap-14 and Gap-01.
(a) Precise statement. A defined partition of field content into system (BH) / bath (radiation) with a consistent pointer basis and reduced dynamics. Gap-14's CH-2 decoherence channel currently honest-halts (NotImplementedError), cascading on Gap-01.
(b) Why it's hard / prior-attempt lessons. This is owner-physics on the owner's machine (the recorded next action is to re-examine the system–bath split on the existing 13D fields first, before any rule-layer extension). Not AI-attemptable. Trap: do not treat the decoherence NotImplementedError as a closure-blocker that the agent can route around — it is an honest halt to be discharged by the owner.
(c) Exactly what closes it. Exhibit a structurally consistent pointer basis on the 13D fields and reduced dynamics — OR a refutation (an inconsistent pointer basis is a decision-grade falsifier).
(d) Machinery & inputs. Open-quantum-systems / decoherence machinery on the frozen 13D field content; Gap-14's CH-2 channel. Owner's machine. See TOE_FINAL_merged.html Gap-14 capsule.
(e) Leverage. Unblocks H4 (the Page mechanism). It is Gap-14's deliverable consumed here.
(a) Precise statement. Produce or surface the re-runnable area-law integral that yields $S = A/4G$ at 0.0028% relative error, freeze-before-compare clean: the effective action, graviton/zero-mode reduction, $G$ normalization, and comparison target. Currently lives only in the external Gap-13 dossier; not reproduced in the rendered corpus.
(b) Why it's hard / prior-attempt lessons. It is finite and setup-able once the action and reduction are handed over; the difficulty is purely that the witness is not yet in the rendered corpus. Trap: do not reconstruct or fabricate the 0.0028% number — surface the actual computation; a fabricated value is worse than an honestly-marked OPEN audit item.
(c) Exactly what closes it. A re-runnable script/record reproducing the value-match. Its absence weakens only the diagnostic's auditability, not the gate's openness.
(d) Machinery & inputs. The external Gap-13 dossier's area-law derivation; the frozen zero-mode reduction and $G$ normalization. Audit-only.
(e) Leverage. Audit hygiene; strengthens the banked diagnostic's credibility. No physics leverage on the count or mechanism.
(a) Precise statement. (i) The internal geometry ($K_6 = SU(3)/T^2$, $S^2$, $S^1_Y/\mathbb{Z}_2$; spin-$\mathbb{C}$ index $= -3$) carries topological invariants whose rank is reserved (PC-4) as a candidate microstate reservoir; the map from rank to a degeneracy reproducing $A/4G$ is absent, and the handle is held deliberately inert. (ii) A pre-registered, mechanical discharge-gate registry (MO-13-6): the condition under which each held non-claim (PC-1, PC-3, PC-4) may be lifted. (iii) Confirm the clause-wise PC-1…PC-5 reconstruction matches the dossier's verbatim disclaimer texts (the wall's coverage was verified against a reconstruction, not the verbatim texts which live external).
(b) Why it's hard / prior-attempt lessons. (i) is doubly gated (needs the UV Hilbert space and the activation gate); until both, PC-4 forbids any state estimate from the rank — do not invoke it early. (ii) and (iii) are governance/documentation, not physics — the agent refuses to invent the registry (fail-closed); these route to the owner (TOE-GAP-13-Q06/Q07).
(c) Exactly what closes it. (i) Pass the pre-registered activation gate (prove the rank may legitimately be the state source) and derive the rank → degeneracy map. (ii) Owner supplies the governance registry pre-registering each PC clause's discharge condition. (iii) A finite documentation check confirming verbatim PC force matches the reconstruction; if it differs, coverage must be re-checked.
(d) Machinery & inputs. (i) Topological/index methods on the frozen internal geometry; the activation gate. (ii)/(iii) governance docs and the external dossier's verbatim PC texts.
(e) Leverage. (i) Would supply an alternative state reservoir for the count (parallel to H3). (ii)/(iii) keep the wall honest and the future claim gated, not vibes.
Dissolved ≠ solved. The cost-floor / granularity axiom dissolves the continuum near-horizon UV-completion prerequisite (a fixed resolution $\ell_* \sim \Lambda_{\rm YM}^{-1}$ declines the $a\to 0$ idealization at the horizon). This removes a prerequisite burden — it does not compute the microstate count, does not supply $1/4$, and does not lift the gate. The printed framework verdict for the Gap-13 whole is untouched.
Selection ≠ derivation; given-E ≠ derivation-of-E. The geometry-essential content (the perturbative graviton zero-mode reduction) is real, but it is precisely not where the gate lives (the finite-area horizon). It terminates on E (the linearized-GR / Einstein–Hilbert limit, carried as given-E input) plus the measured Newton constant $G$. The 0.0028% diagnostic rides on that inherited sector — so the carrier is an inherited route statement, not a banked geometry-output recovery, and the graviton gate it rides on is itself CERTIFICATE-CONDITIONAL / OPEN.
ANCHORED ≠ DERIVED. Even in the best case (T2), the coefficient $1/4$ is anchor-transferred ($1/4 \to G \to$ carried posits + one unproven theorem), never anchor-eliminated. It terminates on posits, not on a derivation. The floor is $\geq 1$ forever; no new tuning was introduced or removed.
What is explicitly NOT claimed. No derived coefficient $1/4$. No horizon microstate count. No quantitative Page curve (only a qualitative, area-dominated turnover, not a theorem of unitarity). No closed gate. No status upgrade — STATUS-UPGRADES:0.
The dissolved unicorns (shared ceilings, never claimed as proven, never listed as our weakness). - "No simpler / no horizon-capable formulation could ever exist for this geometry" — a universal negative over an open-ended space of future formulations, unprovable in principle for any theory in any field. The bounded claim (a named, gated route to a count + a mechanism) is the ceiling, not a hedge. - "THE unique universal microstate count valid for any realistic gravity theory" — demanding a universal-over-all-theories result is unprovable for everyone; the honest object is a geometry-native counting theorem for THIS geometry (bounded; S2.c). - "No future theory could derive the Page curve better" — a forward-looking universal negative over all possible future physics; dissolve, do not chase. The bounded object is whether THIS geometry's $S_{\rm rad}(t)$ turns over — testable, listed as a real hole (H4).
The anchors paid. Two carried posits (AX-INDUCED-G + AX-ENT-EXHAUSTS, the latter replaced in the coefficient argument by the value-free AX-SADDLE-ENTROPY), over a generic unitarity floor, plus the separately-carried open-premise finite-dimensionality posit and an inherited regulator (zero new cost), plus the measured Newton constant $G$ as the IR anchor. Floor $\geq 1$.
Honest ceiling, stated plainly. Gap-13 is a serious candidate / partial unification — NOT validated: an honestly-held OPEN gate that has been sharpened and adjudicated, not closed. The genuine gains are real — a gating theorem run to a refutation that pins the irreducible seam; a coefficient $1/4$ moved from "tautological / zero evidential weight" to "non-tautological, DERIVED-GIVEN-E conditional on one named, well-posed, unproven theorem"; an explicit island/QES construction that locates the island at the textbook position from frozen inputs; a banked area-law value diagnostic at 0.0028%; and a clean firewall against confinement/horizon conflation. None of these is a promotion. The gate cannot move to "Reduced to Axiom," because the closing posit would be true-by-construction. The correct disposition is OPEN.
Dossier synthesized 2026-06-29 from the Gap-13 completion-handoff packet (PER_GATE_DOSSIERS/GAP13_COMPLETION_HANDOFF/: 01_DOSSIER.md, 02_CURRENT_STATE.md, specialist_reply.txt, 08_FOUND_AXIOMS_AND_BUILD_PLAN.md, 09_HORIZON_ADMISSIBILITY_PASS_2026-06-25.md, T2_CONICAL_DEFECT_CERTIFICATE.md, certificates/G13_T2_horizon_EH_replica_saddle/CERTIFICATE.md, 07_OWNER_HINTS_AND_DIRECTIONS.md), the canonical capsule (TOE_FINAL_merged.html Gap-13 section), the missing-object inventory (TOE_SELF_CLOSE_OPEN_ITEM_INVENTORY.md), the cost-floor ledger (COST_FLOOR_WALL_IMPACT_LEDGER.md), and the existing brief/result articles (GATE_BRIEF_GAP13.md, GAP13_COMPLETION_RESULT.md). Frozen branch dcc66f1b2685 / a5b1e6f9d951 READ-ONLY. STATUS-UPGRADES:0. The trap is anchoring on the target.