# SG-6 — Moduli Stabilization (GUT Gate 6): Per-Gate Closure-Attack Dossier

> **What this is.** The closure-attack packet for **SG-6 (Moduli stabilization / Gate 6)** on the **frozen
> 13D K₆ branch ONLY**. It recaps how our geometry controls the downstream-used moduli (concise; the full
> derivation lives on the published site), verifies where the status actually stands, names every open
> residual, and lays out the attack plan to push the gate further. It is **not** a rival comparison (that is
> `BATTLE_GATES/`), **not** the one-line status ledger, **not** a reprint of the manuscript.
>
> **Binding discipline (carry verbatim).** No status was ever upgraded. Frozen branch `dcc66f1b2685` / `a5b1e6f9d951`
> READ-ONLY. Honest throughout: the gate certifies **admissibility-restriction + LOCAL stability at the
> chamber-center witness + phenomenological sufficiency** for Gates 1–10 — **global stabilization is an
> explicit NON-claim** (F.9.4 / F.10.1); the **positive-definite moduli-mass Hessian is DIAGNOSTIC-ONLY**
> (F.9.5, the named most-natural attack surface); witnesses fix the moduli **WITHIN the already-selected
> chamber** (given-E, given-the-selected-geometry). The acute soft spot is **θ_H⋆ ≈ 2.46×10⁻¹⁴** (the
> Hosotani phase), **READ from the chamber one-loop V_Hos minimum**, carrying **~85% of the electroweak
> hierarchy** via v_EW = θ_H⋆/(2πR_γ) — a value extracted from a minimum, not independently derived.
> Closure paths must be **non-target-loaded** (the κ³/π kill-test: a proposed axiom counts only if it would
> be written WITHOUT knowing the target). **given-E ≠ derivation of E.** AXIOM-CLOSED ≠ proven; selection ≠
> derivation; dissolved ≠ solved. **The broader four-term V(σ) discharge lives only in downstream Paper IV
> (EXTERNAL; its own caveats) and changes NOTHING in this gate's certificate.**

---

## 0. Header & Verdict

| Field | Value |
|---|---|
| **Gate id** | SG-6 — Moduli stabilization (the Weyl-rigid chamber-center witness) |
| **GUT manuscript gate** | Gate 6 (§6.6; narrative §5.5; certificate `certificates/G06_stabilization/`; authority **Appendix F**) |
| **Status label (binding)** | **PARTIAL** |
| **Manuscript card status** | *Claimed certificate pass under declared admissibility and moduli-control assumptions* (the manuscript's own conditional phrasing; PARTIAL is the honest scoped-GUT roll-up of it) |
| **Frozen hashes it rides** | Branch `dcc66f1b2685` / manifest meta `a5b1e6f9d951`. Moduli primitives: three Cartan radii of $K_6$ R1.2 `634438ce0776` / `2381d472c62e` / `0e8b8dba2cf0`; chamber-center $\vec u=(1,1,1)$ (A1.2); modular fixed point $\tau=\omega$ `03b30a9c931a`; Higgs winding $n_H=1$ + cycle $\gamma$ (A1.12); finite chamber determinant $\eta_{BK}=0.009721281516312$ `84e94518d3f5`; RG-transport rule `f531205a9159`; comparison scale $M_Z$ `a6852c7a6b00`; threshold spectrum $(\delta_1,\delta_2,\delta_3)=(+4.8424,-3.1112,-1.7313)$; species normalizations $N_{u,d,e}$ `20dc4e0b8220`; chamber angle $\theta_F$ `1ff57f48d45a`; spin-ℂ index / family count $\chi(K_6,E)=-3$ (R1.4) `0fd19c9ae0c1`. **θ_H⋆ (Hosotani minimum) and $M_R$: no SG-6 hash — read/uncomputed.** |
| **Target anchor(s)** | **No new anchor.** Witness ledger terminates on **spectrum-E** (given-E, AXIOM-CLOSED) and the **ℏ-footing only** (μ_cell, SCHEME-ANCHORED/BLOCKED); **global stabilization = explicit NON-claim** (terminates on no number). The one acute open leg, **θ_H⋆** (read from the V_Hos minimum, ~85% of the EW hierarchy), must reproduce **v_EW = 246 GeV** — here **measured-but-irreducible**; θ_H⋆-as-derived stays **OPEN**. Hardening move = the **θ_H⋆ / Hessian derivation**, not a new datum. See §A target-anchor block. No status was ever upgraded. |

### 0.1 Abstract — established / open / what would close it

**Established (given-E, given the selected & frozen chamber).** Gate 6 is the *curve-not-a-number* discipline
made operational (§6.6 "$y=R^2$" toy): **every modulus an output reads must carry a structural witness of a
declared TYPE** — Weyl-rigid chamber for $K_6$'s three Cartan shape moduli (chamber-center $\vec u=(1,1,1)$),
modular fixed point $\tau=\omega$ for the $F^+$ Cartan-torus, integer winding $n_H=1$ for the Wilson-line
Higgs, discrete $\mathbb{Z}_2/\mathbb{Z}_6$ topology for the orbifold/identification, plus
boundary-condition pins for the $S^2/S_Y^1$ radii and the threshold spectrum — **not a scalar potential whose
minimum was placed at the answer.** The witness-type **predicate** tests the *type* of the witness, so "a
potential we minimized at $R_0$" **fails by construction**. The deliverables are the **12-row witness ledger
(F.3)**, a **no-tachyon check (F.4)**, a **witness-coverage lint**, and the five-concept claim-type ledger
(F.9.1 / F.10.1) that explicitly separates what is proved from what is not. This is genuinely strong
discipline and is the program's honest answer to the "hidden-knob" attack.

**Open (the honest deductions).** (1) **Global stabilization is an explicit NON-claim** (F.9.4 row 1 /
F.10.1) — outside-chamber branches are *rejected by admissibility, not stabilized*. (2) The **positive-definite
moduli-mass Hessian is DIAGNOSTIC-ONLY** (F.9.5); a precise positive-definite Hessian is "out-of-scope
refinement" — and the **named most-natural attack surface** is F.9.7(2): *show the chamber-center witness is a
saddle, not a minimum*. (3) The acute soft spot is **θ_H⋆ ≈ 2.46×10⁻¹⁴** — the Hosotani-phase minimum that
carries **~85% of the EW hierarchy** via $v_{\rm EW}=\theta_H^\star/(2\pi R_\gamma)$ — **read from a one-loop
$V_{\rm Hos}$ minimum, not derived-small** (frozen $\eta_{BK}=0.009721$ gives only $\sqrt{\eta_{BK}}/2\pi\sim
10^{-2}$, ~12 orders short). (4) The witnesses fix the moduli **WITHIN the already-selected chamber**
(given-E): the gate certifies the selected geometry is internally controlled, **not** that the witnesses
determine the geometry. (5) Stabilization is a **SHARED-OPEN** hard problem — rivals (string/M/F/NCG/lattice)
also struggle (Battle verdict: TIE/SHARED-OPEN, **not** a Fable win).

**What would close it (the ladder).** DERIVED-CLOSED would require either (a) the **four-term V(σ) discharge**
turning the breathing/volume modulus from witness into a *derived* positive-Hessian minimum with no new knob,
or (b) a **target-blind derivation of θ_H⋆** (the small Hosotani phase from chamber data, not read from
$V_{\rm Hos}$), or (c) a **positive-definite Hessian certificate** at $(1,1,1)$. The realistic ceiling per
residual is **AXIOM-CLOSED**: name one explicit, target-blind posit (the modular-fixed-point uniqueness; the
uniform operational cell value $\Delta_0$ that pins the loop scale; the second-dimensionful-anchor irreducibility
of $v_{\rm EW}$). The honest expected outcome on the two genuine-physics fronts (shape-doublet Hessian, θ_H⋆)
is **sharper-OPEN** — and on the shape doublet, on current Λ-free perturbative geometry, **a SADDLE** (a
falsifier-grade negative, not a closure).

### 0.2 Website source-of-truth (link, don't duplicate)

The full construction, the witness ledger, and the claim-type discipline are the **published manuscript**,
Paper I:

- **GUT.html §6.6** (Gate-6 card) — <https://physics.magflowmeters.com/articles/GUT.html>
- **§5.5** narrative module; the **$y=R^2$ curve-not-a-number** discipline; **Appendix F** (Stabilization —
  the binding authority): **F.1/F.2** witness structure, **F.3** the 12-row moduli ledger, **F.4** tachyon
  check, **F.5** sensitivity, **F.9** Stabilization-vs-Admissibility five-concept distinction (**F.9.4**
  NON-claim, **F.9.5** Hessian diagnostic, **F.9.7** falsification paths), **F.10** claim-type ledger +
  required-vocabulary substitution rule; **Appendix CR module CR6** reader companion. Same public article.
- Downstream cross-reference (does **not** touch this gate's certificate): the four-term $V(\sigma)$ discharge
  in **Paper IV, TOE.html** — <https://physics.magflowmeters.com/articles/TOE.html> (EXTERNAL; its own
  caveats). The θ_H⋆ / $v_{\rm EW}$ leg is the Gate-8 Hosotani object (§6.8 / Appendix H) on which SG-6's
  acute soft spot rides.

This dossier recaps only what is needed to attack; the site controls all common material.

---

## 1. How OUR geometry controls the moduli — the closure claim (concise)

> Full derivation: **GUT.html §6.6 + §5.5 + Appendix F (F.1–F.10) + CR6**. This section is the attack-grade
> recap, not the derivation.

### 1.1 The mechanism end-to-end

Gate 6 answers a single plain fact: **a predictive compactification cannot have its hidden shape and size
drifting freely — every modulus any later gate reads must be pinned before a number is compared** (§5.5; the
$y=R^2$ toy, §6 "Curve-not-a-number"). The Fable mechanism is **not** "write a scalar potential, minimize it,
read off the value." It is a **witness-type predicate**: each downstream-used modulus must carry a witness
drawn from a declared set of *structural* mechanisms, and the predicate tests the **type** of the witness, so
a tuned potential whose minimum was *placed* at the answer fails by construction (R1.6 forbids exactly this).

The load-bearing factors, named so a reader can attack each:

1. **The Weyl-rigid $K_6$ chamber** (the three Cartan shape moduli $\vec u=(u_1,u_2,u_3)$ of
   $K_6=SU(3)/T^2$). Witness type = **admissibility / no-runaway on the spin-ℂ bundle data**: off-chamber
   moduli fail the A0 admissibility check and the branch is *eliminated by the selector*
   (Appendix B1.6), not stabilized. The chamber is $\vec u\in[1/2,3/2]^3$ with center pinned at $(1,1,1)$.
   *(Attack handle: the Weyl group $S_3$ permutes $(u_1,u_2,u_3)$, so $(1,1,1)$ is the $S_3$-fixed point —
   automatically a **critical point** of any $S_3$-invariant functional, but the **Hessian sign** is the open
   object; see §3 R3.)*
2. **The modular fixed point $\tau=\omega=e^{2\pi i/3}$** (`03b30a9c931a`) for the $F^+$ Cartan-torus modulus.
   Witness type = **residual modular symmetry forces the order-three fixed point**; small perturbations off
   $\omega$ acquire a non-zero potential under the chamber's RG transport, restoring $\tau\to\omega$.
   *(Attack handle: this is the cleanest witness — a symmetry-protected fixed point, not a minimum read off a
   curve. But the manuscript's own F.2 wording ("produce a non-zero potential off the fixed point") leans on a
   potential it does not compute; the symmetry-uniqueness leg is the strengthening target, R7.)*
3. **Integer Wilson-line winding $n_H=1$** on $K_{\rm gauge}$, on the declared homology cycle $\gamma$.
   Witness type = **topological invariant** — $\delta n_H\notin\mathbb{Z}$ is forbidden; integer winding is the
   only allowed value. *(Attack handle: the winding **count** is protected, but the **phase** $\theta_H$ that
   the Wilson line settles at is a continuous modulus on the cycle — and **that is θ_H⋆, the acute soft spot**;
   see §3 R1.)*
4. **Discrete $\mathbb{Z}_2$ / $\mathbb{Z}_6$ topology.** The $S_Y^1/\mathbb{Z}_2$ orbifold action $y\mapsto-y$
   and the global $\mathbb{Z}_6$ identification have **no continuous deformation parameter to drift** — fixed
   by topology, no witness-potential needed.
5. **Boundary-condition pins for the $S^2/S_Y^1$ radii and the threshold spectrum.** $R_{S^2}$ by KK /
   Wilson-line balance against the threshold-unification target $M_U$; $R_{S_Y^1}$ by the hypercharge
   gauge-coupling boundary condition at $M_Z$; the threshold vector
   $(\delta_1,\delta_2,\delta_3)=(+4.8424,-3.1112,-1.7313)$ from the KK tower. *(Attack handle: these are
   **threshold-pinned**, i.e. their witness is consistency with Gate 7 — they inherit Gate 7's open status,
   notably the δ-triple reproducibility gap; given-E, given the frozen scheme.)*
6. **The 12-row witness ledger (F.3) + no-tachyon check (F.4) + coverage lint.** Every modulus / Wilson-line /
   chamber coordinate / threshold spectrum that any Gate 1–10 reads is listed with its witness, pinned value,
   downstream consumer, and the effect of a small perturbation $\delta$. The certificate JSON (F.6) reports
   `stabilized_object_count: 6` "potential/witness" classes covering the 12 ledger rows, `tachyon_check: no
   tachyon on the active branch`.

The pipeline is one discipline, not one computation:

```
 every downstream-used modulus
   --witness-type predicate-->  {Weyl-rigid | modular fixed pt | integer winding | discrete topology | BC pin}
     --type test (NOT minimize-at-answer)-->  PASS (witnessed) / FAIL (tuned-or-floating)
       --12-row ledger + no-tachyon + coverage lint-->  Claimed certificate pass (given-E, in-chamber)
```

### 1.2 What "closed" means here, and its conditionality

"Closed" for SG-6 is **moduli-control by structural witness under declared assumptions** — strictly **not**
global stabilization, **not** a proven positive-definite Hessian, **not** a derivation of the chamber. It is
the conjunction of exactly three claim types (F.9.3 / F.10.3):

- **(i) Admissibility restriction** — non-chamber configurations rejected by $\mathcal{C}_{\rm admiss}$
  (Appendix C6). *Proven* (it is an $\oplus$-layer rulebook rule).
- **(ii) Local stability at the chamber-center witness $\vec u=(1,1,1)$** — small in-chamber perturbations
  return to the center. *Claimed* at tree-plus-one-loop within the declared scheme.
- **(iii) Phenomenological sufficiency** — the remaining moduli do not move any Gate 1–10 output beyond its
  published tolerance. *Claimed* under the Weyl-rigid chamber + freeze discipline.

It is conditional on:

- **given-E** — the SM chiral content and the family index $\chi=-3$ are supplied as input from SG-2/SG-3; the
  witnesses operate *on the selected bundle*, they do not derive it.
- **given-the-selected-geometry** — the witnesses fix the moduli **inside** the already-selected chamber. SG-6
  certifies the selected geometry is internally controlled; it does **not** prove the witnesses determine the
  geometry. *(This is the load-bearing conditionality of the whole gate.)*

**What the gate explicitly does NOT claim** (F.9.4 / F.10.1, carried verbatim):

| Concept (F.9.1) | Claimed? | Status |
|---|---|---|
| Admissibility restriction | **Yes** | proven (rulebook) |
| Local stability at $(1,1,1)$ | **Yes** | claimed, tree+1-loop, declared scheme |
| Moduli mass generation (positive-definite Hessian of $V(\vec u)$) | **Partial → Diagnostic only** | F.9.5; **the named attack surface (F.9.7-2)** |
| Global stabilization (no flat directions anywhere) | **NOT claimed** | F.9.4 row 1 explicit non-claim |
| Phenomenological sufficiency for Gates 1–10 | **Yes** | claimed |

So the precise statement of closure: **the gate certifies that every downstream-used modulus carries a
structural witness (not a tuned potential), that small in-chamber perturbations are controlled at the
chamber-center, and that the residual moduli do not spoil Gates 1–10 — given the SM content and the selected
geometry. It does NOT prove a positive-definite Hessian, does NOT claim global stabilization, and the one
modulus carrying the electroweak hierarchy (θ_H⋆) is READ from a minimum, not derived.**

---

## 2. Verify the status — is PARTIAL real?

This is the verification a skeptic would run. Each witness gets a grade — **hand-checkable** / **symbolic** /
**machine-lane** — and an honest **reproduces?** flag.

### 2.1 The witness ledger

| # | Witness | What it asserts | Grade | Reproduces? |
|---|---|---|---|---|
| W1 | **Witness-type predicate** (R1.6: "no tuned potential whose minimum was chosen at the answer") | every downstream modulus carries a witness of declared TYPE; "minimize-at-answer" fails by construction | symbolic/audit | **Yes** — the predicate is a type-check, hand-auditable against F.3; it is the strongest, most defensible leg |
| W2 | **12-row moduli ledger (F.3)** | every modulus/Wilson-line/chamber coord/threshold spectrum used by Gates 1–10 is listed with witness + pinned value + downstream consumer + $\delta$-response | hand-checkable | **Yes** — the table is complete and self-consistent; coverage lint passes (no floating downstream modulus) |
| W3 | **Weyl-rigid chamber + center $\vec u=(1,1,1)$** | admissibility rejects off-chamber; center is the $S_3$-fixed point | hand-checkable | **Yes for admissibility + critical-point**; the *Hessian sign* is NOT asserted here (it is the diagnostic) |
| W4 | **Modular fixed point $\tau=\omega$** | residual modular symmetry forces the order-three fixed point | hand-checkable | **Yes (as a fixed point)**; the "non-zero potential off $\omega$" claim is symbolic, not computed |
| W5 | **Integer winding $n_H=1$** | topological invariant; non-integer forbidden | hand-checkable | **Yes** — pure topology |
| W6 | **Discrete $\mathbb{Z}_2/\mathbb{Z}_6$** | no continuous deformation parameter | hand-checkable | **Yes** — pure topology |
| W7 | **No-tachyon check (F.4)** | no tachyonic mode on the active branch at the declared witnesses | symbolic | **Conditional** — asserted for the declared spectrum; a deformation away from a witness *could* produce one (then the certificate fails) |
| W8 | **Phenomenological sufficiency** | residual moduli leave Gates 1–10 within tolerance | machine-lane | **AUDIT** — depends on the same per-gate harnesses (e.g. the Gate-7 threshold pipeline) that are themselves AUDIT |
| W9 | **Positive-definite Hessian of $V(\vec u)$** | $(1,1,1)$ is a minimum, not a saddle | machine-lane | **DIAGNOSTIC-ONLY (F.9.5)** — *not* computed in-gate; the named attack surface. **An out-of-gate Λ-free computation gives a SADDLE in the shape sector (see §3 R3).** |

### 2.2 What reproduces, plainly

- **The witness-type discipline reproduces by inspection.** The predicate is a *type* test; a reader can walk
  F.3 row by row and confirm each witness is structural (admissibility / symmetry fixed-point / integer
  winding / discrete topology / BC pin) and not "a number we minimized to." This is the genuine, strong leg of
  the gate, and it is exactly the $y=R^2$ curve-not-a-number toy applied to every dial the certificates touch.
- **The discrete/topological witnesses reproduce by hand.** $n_H=1$, the $\mathbb{Z}_2$ orbifold, the
  $\mathbb{Z}_6$ identification, the cycle $\gamma$ — all are topological invariants with no continuous knob;
  nothing to reproduce numerically and nothing to drift.
- **The $S_3$-fixed-point structure reproduces by hand.** Because the Weyl group $S_3$ permutes
  $(u_1,u_2,u_3)$, the center $(1,1,1)$ is automatically a **critical point** of any $S_3$-invariant
  functional ($\partial V=0$ for free). This upgrades the manuscript's "fixed by fiat" wording to "fixed by
  symmetry **as a critical point**" — a genuine, free, target-blind structural fact (TEST 2).

### 2.3 What does NOT yet reproduce (honest gaps in the status)

- **The positive-definite Hessian does NOT reproduce — it is diagnostic-only, and the one Λ-free piece that
  CAN be computed gives a SADDLE.** F.9.5 explicitly declines the Hessian computation as "out-of-scope
  refinement." When the chamber Hessian is decomposed by symmetry into $3=\mathbf{1}\oplus\mathbf{2}$ (TEST 2),
  the **shape doublet** ($S_3$-doublet, traceless) is computable Λ-free from the flag-manifold scalar
  curvature — and the result is $\partial^2 R/\partial\varepsilon^2|_{(1,1,1)}=-1<0$, **a saddle** (the
  structure-constant triangle term $-5/2$ overwhelms the diagonal $+2$). The **breathing/volume singlet** rides
  the uncomputed $c_{\rm loop}$ (the 13D FRG-2 wall). So the F.9.7(2) saddle attack surface is not merely
  *open* — on current perturbative geometry it **leans toward saddle**, rescuable only by a Casimir term whose
  net sign needs the absent admissible-rep multiplicity table (TEST_CASIMIR_RESCUE).
- **θ_H⋆ does NOT reproduce as a derivation — it is read from a minimum.** The Hosotani phase
  $\theta_H^\star\approx2.46\times10^{-14}$, which carries ~85% of the EW hierarchy via
  $v_{\rm EW}=\theta_H^\star/(2\pi R_\gamma)$, is **extracted from the one-loop $V_{\rm Hos}$ minimum**, not
  computed-small from chamber data. The frozen $\eta_{BK}=0.009721$ gives only $\sqrt{\eta_{BK}}/2\pi\sim
  10^{-2}$ — ~12 orders short. The lightness-of-$v$ leg leans on a read modulus.
- **Global stabilization does NOT reproduce — it is not claimed.** F.9.4 row 1 says so explicitly;
  outside-chamber branches are *rejected by admissibility, not stabilized*. There is no all-loop
  positive-definite Hessian and no claim of "no flat directions anywhere."
- **Phenomenological sufficiency inherits downstream AUDIT.** The "residual moduli don't spoil Gates 1–10"
  claim is only as machine-real as the downstream harnesses it relies on — and the Gate-7 threshold pipeline
  (which consumes the SG-6 radii/spectrum) is itself AUDIT (δ-triple injected reals, `reproduce_all.py`
  absent). SG-6's sufficiency claim is therefore declared-not-independently-verified at the numerical level.

### 2.4 Why PARTIAL (not higher, not lower)

- **Not DERIVED-GIVEN-E** (the SG-2/SG-3 tier): SG-6 is not a rigid integer/representation recovery. It has a
  diagnostic-only Hessian (with a Λ-free saddle in the shape sector), a read-from-minimum θ_H⋆ carrying the
  hierarchy, an explicit global-stabilization non-claim, and downstream-inherited AUDIT — too many open
  residuals for the DERIVED tier.
- **Not OPEN/DECLARED-FROZEN:** the witness-type discipline is real, strong, and non-trivial; the
  discrete/topological witnesses and the $S_3$-critical-point structure are genuine target-blind facts; the
  claim-type ledger honestly fences what is and is not proved. The gate is well clear of "unfixed knobs."
- **PARTIAL is exactly right:** genuine structural moduli-control under declared assumptions, a diagnostic-only
  Hessian named as the most-natural attack surface, an explicit non-claim on global stabilization, and a
  read-from-minimum acute soft spot. This matches the SCOPED_GUT ledger SG-6 line verbatim.

---

## 3. The attack surface — what to attack (ranked by leverage)

Every open residual as a named object with its precise obstruction. **Leverage** = how much closing it moves
the gate (and, for SG-6, how much of the electroweak hierarchy / how much of the certificate it touches).

### 3.1 The residual register

| ID | Residual (named object) | Precise obstruction | Status | Leverage |
|---|---|---|---|---|
| **R1** | **θ_H⋆ ≈ 2.46×10⁻¹⁴ read from the $V_{\rm Hos}$ minimum** | The Hosotani phase carries **~85% of the EW hierarchy** ($v_{\rm EW}=\theta_H^\star/(2\pi R_\gamma)$) yet is **read from a one-loop minimum, not derived-small**; $\sqrt{\eta_{BK}}/2\pi\sim10^{-2}$ is ~12 orders short. The lightness-of-$v$ leg leans on a read modulus. | OPEN (acute soft spot) | **HIGHEST** — it carries the bulk of the hierarchy and is the gate's single most attackable physics object |
| **R2** | **Global stabilization is an explicit NON-claim** | F.9.4 row 1 / F.10.1: outside-chamber configurations are *rejected by admissibility, not stabilized*; no all-loop positive-definite Hessian; no "no flat directions anywhere." | DISCLOSED (explicit non-claim) | **MEDIUM** — honesty/claim-boundary; the SHARED-OPEN hard problem; closing it is a known no-go-adjacent program |
| **R3** | **Positive-definite Hessian is DIAGNOSTIC-ONLY (the named attack surface)** | F.9.5 / F.9.7(2): "show the chamber-center witness is a saddle, not a minimum." The Λ-free shape-doublet Hessian = **−1 (SADDLE)** on perturbative geometry; the breathing singlet rides the uncomputed $c_{\rm loop}$. | OPEN — **saddle-leaning** | **HIGH** — this is the explicitly-named falsification path; the shape sector is *already* computed to a saddle |
| **R4** | **Witnesses fix moduli WITHIN the selected chamber (given-E)** | The gate certifies internal control of the *selected* geometry; it does not prove the witnesses *determine* the geometry. Selection ≠ derivation; given-E ≠ derivation of E. | OPEN (inherited from SG-1/SG-3) | **MEDIUM** — conditions every SG-6 claim; shared with SG-1 |
| **R5** | **$c_{\rm loop}$ / breathing-mode singlet = the 13D FRG-2 β-vector wall** | The volume/breathing direction's Hessian sign rides the uncomputed $c_{\rm loop}$ (the $e^{-6\sigma}$ coefficient = $a_6$ of the σ-fluctuation determinant). $c_{\rm loop}$ is `BLOCKED_MISSING_FRG_MATCHING`; even a finite cell-sum leaves one log-scheme factor $\mu_{\rm cell}$ free. | OPEN / BLOCKED_INPUTS | **MEDIUM-HIGH** — shared deep wall with Gap-04 / hierarchy / a6; closing it decides the singlet Hessian sign |
| **R6** | **Casimir net sign needs the absent admissible-rep multiplicity table** | The shape-doublet saddle can be rescued by a graded Casimir term; per-sector sign is LOCKED (bosons stabilize, fermions destabilize), but the **net sign** needs the multiplicity table (ABSENT) and a magnitude $\mu_{\rm cell}\cdot q\gtrsim0.5$. | BLOCKED_INPUTS / NEEDS-TUNING | **MEDIUM** — the only route that could flip R3 from saddle to minimum; gated on a missing owner artifact |
| **R7** | **$\tau=\omega$ "non-zero potential off the fixed point" leans on an uncomputed potential** | F.2 asserts perturbations off $\omega$ "produce a non-zero potential," but the potential is not computed; the leg should rest on **modular-symmetry uniqueness** of $\omega$, which is not proven in-gate. | OPEN | **LOW-MEDIUM** — the cleanest witness; strengthening it from "potential" to "symmetry-protected" is a contained group-theory win |
| **R8** | **No-tachyon check is conditional on the declared spectrum** | F.4: no tachyon "on the active branch"; a deformation away from a witness could produce one. The check is asserted for the frozen spectrum, not proven stable under the deformation classes the gate exposes. | DISCLOSED (conditional) | **LOW** — disclosed; the deformation classes are bounded by admissibility |
| **R9** | **Phenomenological sufficiency inherits downstream AUDIT (Gate-7 radii/spectrum)** | The "residual moduli don't spoil Gates 1–10" claim relies on the Gate-7 threshold pipeline (which consumes the SG-6 radii + threshold spectrum), which is itself AUDIT (δ-triple injected, `reproduce_all.py` absent). | AUDIT (inherited) | **LOW-MEDIUM** — required for the sufficiency leg to be machine-real; the artifact is owned by SG-7 |

### 3.2 Leverage ranking (attack order)

1. **R1** (θ_H⋆ read from a minimum) — carries ~85% of the hierarchy; the single highest-value physics target.
2. **R3** (the diagnostic Hessian — the named saddle attack surface) — already computed to a saddle in the
   shape sector; the explicit F.9.7 falsification path.
3. **R5** (the $c_{\rm loop}$ / breathing-mode singlet wall) — decides the singlet Hessian sign; the shared
   deep 13D FRG-2 wall.
4. **R6** (the Casimir rescue) — the only route that could flip R3; gated on the missing multiplicity table.
5. **R4** (given-E / within-chamber) — conditions everything; honesty residual shared with SG-1/SG-3.
6. **R2 / R7 / R8 / R9** — disclosed non-claim / cleanest-witness strengthening / conditional tachyon /
   inherited AUDIT.

> **The cardinal honest point.** R1 and R5 are where the "85% hierarchy" and the "singlet Hessian" live, and
> **both reduce to the same single object: $\mu_{\rm cell}$, the spectral value of the uniform operational
> cell $\Delta_0$** (the log-scheme / cell-scale factor of the 6-D $a_6$ determinant). The decisive,
> already-run firewall verdict (CLOSURE_CAMPAIGN §3) is that **$\mu_{\rm cell}$ has no $v$-independent
> readout** — its only available anchor is $\partial_\sigma V=0$, which **IS** the hierarchy — so anchoring
> $\mu_{\rm cell}$ there to "predict" $v$ is **circular by construction (the κ³/π signature)**. Any closure
> path that pins $\mu_{\rm cell}$ from the EW hierarchy has **RELOCATED**, not removed, the input. The κ³/π
> kill-test applies in full force here. R2/R7/R8/R9 are the *non-physics-flip* residuals (claim-boundary /
> symmetry-strengthening / conditional-check / inherited reproducibility); closing them improves honesty and
> machine-reality but **does not derive the hierarchy or flip the Hessian**.

---

## 4. THE ATTACK PLAN — closure paths (the core)

For each residual: the **technique** (closure playbook T1 axiom-floor / T4 no-go / T5 firewall / T6
all-operator-conditional / T7 eliminative / T10 selector), the **named axiom it could reduce to** (stated so it
would be written WITHOUT the target value — the κ³/π kill-test), the **specialist target** (theorem to hand
off) or the **owner artifact** (CSV/ruling/computation) needed, the **math to attempt**, and the **success
ladder** (DERIVED-CLOSED rare → AXIOM-CLOSED likely → sharper-OPEN → REFUTED).

---

### 4.1 R1 — θ_H⋆ ≈ 2.46×10⁻¹⁴ read from the $V_{\rm Hos}$ minimum (the acute soft spot)

**Why first.** This single modulus carries **~85% of the electroweak hierarchy**. If θ_H⋆ could be *derived*
target-blind from chamber data instead of read off the one-loop Hosotani potential, the lightness-of-$v$ leg
would stop leaning on a read number and SG-6 would deliver the hierarchy's dominant factor. If it provably
cannot, the gate's honest standing tightens to "the hierarchy is a second irreducible anchor."

**Technique: T5 (no tuning to the known answer firewall) + T7 (eliminative) — is θ_H⋆ a derivation or a relocation?**
This is exactly the SCALE firewall, pointed at the Hosotani phase.

**The two routes, and which the firewall kills:**

| Route | Statement | Firewall verdict |
|---|---|---|
| **Path A — exact $v$ via $\mu_{\rm cell}$** | derive the loop scale $\mu_{\rm cell}$ from $\{ \hbar, M_U/M_{\rm Pl}, \text{frozen geometry}\}$ at a fixed UV reference, then read $\theta_H^\star$ off $V_{\rm Hos}$ | **CIRCULAR / RELOCATION** — $\mu_{\rm cell}$ must be a *mass*; the only σ-independent mass is $M_{\rm Pl}$ (an anchor), the rest is dimensionless shape; turning shape into a mass needs an inverse length, and the only one available is $R_{K_6}(\sigma)$ — σ-carrying. Buckingham-π: a *second* mass cannot be built from $\{M_{\rm Pl},\hbar,\text{dimensionless geometry}\}$. $\mu_{\rm cell}\to v$ is invertible-by-construction (exponential) ⇒ $\mu_{\rm cell}$ IS the knob. |
| **Path B — structural "huge" via transmutation** | bound $\theta_H^\star$ small by a dimensional-transmutation exponent $t_*=2\pi/(b\alpha)$ | **WRONG MECHANISM** — θ_H⋆ is the **location of a stationary point of a periodic potential**, NOT a running coupling, so there is *no* transmutation exponent to bound. The field-count bound governs only $\ln(M_{\rm Pl}/M_U)$ (the frozen ~15%), never the ~85% in θ_H⋆. (Bonus genuine sub-result: Gate-8 Hosotani protection *does* remove the Higgs-mass quadratic destabilization, but only to ~$10^{14}$ GeV — 12 orders short.) |

**Named axiom it could reduce to (κ³/π-clean).** The honest reduction is **not** a closure but a precise
statement of what R1 reduces to once the two routes are killed:

> **AXIOM-VEW-SECOND-ANCHOR** (target-blind form): *"the electroweak scale $v_{\rm EW}$ is a second
> independent dimensionful anchor on the same footing as $M_{\rm Pl}$, irreducible in this geometry because no
> second mass can be built from $\{M_{\rm Pl}, \hbar, \text{dimensionless frozen geometry}\}$ without a
> σ-carrying length (Buckingham-π), and because the carrier θ_H⋆ is a periodic-minimum location, not a
> transmutation exponent."*

This is writable with **no** value of $v$ or θ_H⋆ in sight — it is a *dimensional + mechanistic* statement
about what kind of object the hierarchy is. **It PASSES the κ³/π kill-test** (it carries no reverse-engineered
coefficient). It does **not** derive θ_H⋆; it names *why* θ_H⋆ must be read, not computed, in this structure.

**Specialist target (to hand off).** A **target-blind θ_H⋆ derivation theorem**: "Given only the frozen
chamber data (cycle $\gamma$, $n_H=1$, $\eta_{BK}$, $R_\gamma$) and a UV reference frozen *before and
independently of* $v_{\rm obs}$, compute $\theta_H^\star$ and check whether it returns
$\approx2.46\times10^{-14}$ with no $v$ in the inputs." If a specialist can produce such a derivation, R1
closes (DERIVED) and the hierarchy's dominant factor is delivered; if the derivation provably requires
$\partial_\sigma V=0$ (the hierarchy itself), R1 is confirmed RELOCATION and AXIOM-VEW-SECOND-ANCHOR stands.

**Math to attempt.** (1) Write $V_{\rm Hos}(\theta_H)$ from the frozen cycle/winding/$\eta_{BK}$ data. (2)
Locate its minimum θ_H⋆ **without** consulting $v_{\rm obs}$. (3) Check whether the minimum lands at
$\sim10^{-14}$ from chamber data alone, or only after a scale is pinned by $v$. (4) Run the
$v$-independence audit: does the threshold-weight swing across the σ-band ($\sim34\times$, the documented
forbidden count-only "35" stationarity) confirm $\mu_{\rm cell}$ is the knob?

**Success ladder.**
- DERIVED-CLOSED: θ_H⋆ computed target-blind from chamber data, returning $\sim10^{-14}$ with no $v$ input →
  the hierarchy's dominant factor delivered. **Unlikely** — the dimensional near-no-go (Buckingham-π) blocks a
  second mass from $\{M_{\rm Pl},\hbar,\text{shape}\}$.
- **AXIOM-CLOSED (likely):** AXIOM-VEW-SECOND-ANCHOR named; θ_H⋆/$v_{\rm EW}$ conceded as a second irreducible
  dimensionful anchor with a *principled* (dimensional + mechanistic) reason, not "we failed to reduce it."
  This is the honest ceiling and is essentially where the SCALE frontier already stands.
- sharper-OPEN: the derivation attempt returns fail-closed on $\mu_{\rm cell}$ (the FRG-2 wall) without
  deciding circularity.
- REFUTED: a chamber-only computation gives a θ_H⋆ that *misses* $10^{-14}$ → the lightness-of-$v$ leg is a
  genuine prediction-vs-data failure.

**Honest disposition: AXIOM-CLOSED at AXIOM-VEW-SECOND-ANCHOR; θ_H⋆ as "derived-small" stays OPEN/RELOCATION.**
The firewall has already confirmed Path A circular and Path B wrong-mechanism; the genuine banked sub-result is
the Higgs-mass quadratic protection (proven, finite, 12-orders-short). Saying precisely "the hierarchy is a
second anchor, here is the structural reason it must be" is the closure — **not** a higher claim.

---

### 4.2 R3 — The diagnostic Hessian: the named saddle attack surface (F.9.7-2)

**Why second.** F.9.5 grades the positive-definite moduli-mass Hessian DIAGNOSTIC-ONLY and F.9.7(2) names the
single most-natural falsifier: *show the chamber-center witness is a saddle, not a minimum.* This is the
explicitly-named attack surface, and it is **already partially executed against the gate.**

**Technique: T7 (eliminative) by symmetry decomposition — split the Hessian, compute the computable piece.**
The Weyl group $S_3$ permutes $(u_1,u_2,u_3)$, so the 3-dim perturbation space splits exactly into
$3=\mathbf{1}\oplus\mathbf{2}$ (TEST 2):

| Mode | Irrep | Stability controlled by | Status |
|---|---|---|---|
| **breathing / volume** ($u_1=u_2=u_3$) | singlet $\mathbf{1}$ | $V''(\sigma)$, the $e^{-6\sigma}$ term → $c_{\rm loop}$ | **= the FRG-2 wall (R5)** |
| **shape** ($u_1-u_2$, ×2 degenerate) | doublet $\mathbf{2}$ | flag-manifold curvature / Casimir | **Λ-free, $c_{\rm loop}$-free, computable** |

$(1,1,1)$ is a minimum **iff both eigenvalues > 0**, and they decouple by symmetry.

**The Λ-free computation, already run (TEST_SHAPE_DOUBLET_STABILITY).** The flag-manifold scalar curvature
$R(u)=\sum_i 1/u_i - \tfrac12 T(u)$, $T(u)=\sum_k u_k/(u_iu_j)$, along the doublet ray
$u=(1+\varepsilon,1-\varepsilon,1)$ gives:
- diagonal $\sum 1/u_i$: $\to +2$ to the $\varepsilon^2$ coefficient (convexity of $1/u$),
- structure-constant triangle $T$: $\to -5/2$,
- net $\partial^2 R/\partial\varepsilon^2|_{(1,1,1)} = +2 - 5/2 = \mathbf{-1} < 0$ → **SADDLE.**

The boundary (Gibbons–Hawking) term is small and mildly destabilizing; the Wilson-line term is exactly zero in
the doublet direction. **On purely perturbative geometry, the shape doublet is a SADDLE** — the
structure-constant coupling between the three root planes makes the equal-scale flag metric unstable to shape
deformation. (This is the well-documented fact that $SU(3)/T^2$ carries three invariant Einstein metrics and
the normal one is not the shape-sector extremizer.)

**The only rescue, and its kill-test (TEST_CASIMIR_RESCUE).** A graded Casimir / KK vacuum-energy term can
flip the doublet sign. The per-sector sign is **LOCKED** (bosons stabilize, fermions destabilize, by
convexity $g''(0)=2(w_1^2+w_2^2)>0$ and $f'(m^2)>0$, scheme-independent). The **net** sign is
$\mathrm{sgn}(\text{boson-surplus}-\text{fermion-surplus})$ over admissible $(p,q)$ at $(1,1,1)$ — which **needs
the admissible-rep multiplicity table (ABSENT from the corpus)** and a magnitude $\mu_{\rm cell}\cdot q\gtrsim
0.5$. The frozen heat-kernel ledger (G.3.2a) *weakly* favors boson dominance (adjoint gauge+ghost
$-2.49/-4.02$ vs matter $+0.79/+0.92$, ~3:1) → conjectured rescue — **but the opposing frozen fact is
$\chi=-3$ (a low-lying fermionic chiral surplus that destabilizes)**, and banking boson-dominance without the
multiplicity table is NEEDS-TUNING.

**Named axiom it could reduce to (κ³/π-clean).** Not a closure, but the precise reduction of the doublet
question:

> **AXIOM-SHAPE-DOUBLET-CASIMIR** (target-blind form): *"the shape-doublet Hessian at the chamber center is
> $-1_{\rm curvature} + q\,\mu_{\rm cell}\,(+2_{\rm convexity})$, with the convexity factor $+2$ and the
> per-sector sign locked by geometry; the net sign is fixed by the graded boson−fermion surplus of the
> admissible-rep spectrum."* Writable with **no** Λ and **no** target value — it references only the
> root-system curvature and the graded multiplicity. PASSES the kill-test as a *statement*; the test is then
> whether the multiplicity table *yields* a net positive.

**Specialist target / owner artifact.** (1) **Owner artifact:** the **admissible-rep multiplicity table** at
$(1,1,1)$ + the regularized zeta (from `certificates/` + `reproduce_all.py` / the KK eigenvalue file R0.7) —
this is the single missing input that decides the net Casimir sign. (2) **Specialist target:** "Given the
multiplicity table, compute $\mathrm{sgn}(\text{boson-surplus}-\text{fermion-surplus})$ and the magnitude
$\mu_{\rm cell}\cdot q$, and report whether the doublet net Hessian is $>0$ (rescue) or $\le 0$ (saddle
persists)."

**Math to attempt.** (1) The doublet curvature $-1$ is **done** (banked, target-blind). (2) Mount the
multiplicity table; form the graded doublet-weighted sum $\sum (-1)^F d\,n\,(w_1^2+w_2^2)$. (3) Decide the net
sign. (4) Combine the singlet (R5) and doublet results for the full Hessian signature.

**Success ladder.**
- DERIVED-CLOSED: both eigenvalues (singlet via R5, doublet via the multiplicity table) proven $>0$
  target-blind → the Hessian is positive-definite, F.9.7(2) defeated, the diagnostic upgrades to a derived
  minimum. **Gated on both the multiplicity table AND the FRG-2 $c_{\rm loop}$ — unlikely in one step.**
- AXIOM-CLOSED: AXIOM-SHAPE-DOUBLET-CASIMIR named and the multiplicity table *yields* a net positive doublet →
  the shape sector reduced to one symmetry/spectral posit. **Best realistic outcome for the doublet.**
- **sharper-OPEN / saddle-leaning (current standing):** doublet curvature is a saddle ($-1$); the rescue needs
  the absent table; the singlet rides the FRG-2 wall. The honest current state is "saddle on perturbative
  geometry, conditional rescue pending owner artifacts."
- REFUTED: multiplicity table computed and the net doublet stays negative → the chamber-center is a **saddle**,
  F.9.7(2) succeeds, and the moduli-mass-generation row of F.9.2 downgrades (per F.10.4-2, to *Open / not
  claimed*).

**Honest disposition: sharper-OPEN, saddle-leaning.** The named falsification path is the *closest to firing*
of any SG-6 residual: the one Λ-free piece that can be computed is a saddle, and the rescue is gated on a
missing multiplicity table and the $\mu_{\rm cell}$ value (shared with R1/R5). **Do not bank the boson-dominance
rescue without the table.**

---

### 4.3 R5 — $c_{\rm loop}$ / breathing-mode singlet = the 13D FRG-2 β-vector wall

**The obstruction.** The breathing/volume singlet eigenvalue rides $V''(\sigma)$ at the $e^{-6\sigma}$ term —
i.e. $c_{\rm loop}$, the coefficient of the σ-fluctuation one-loop determinant, equivalently the **$a_6$
Seeley–DeWitt heat-kernel coefficient** on the 6-manifold $K_6$. $c_{\rm loop}$ is
`BLOCKED_MISSING_FRG_MATCHING` — the uncomputed 13D FRG-2 β-vector, the deep core shared with Gap-04 and a6.

**Technique: T7 (eliminative) — is the wall the continuum or the scale?** Replace the continuum mode integral
by a **finite cell-sum** under the uniform operational cell law $N\le B/\Delta_0$
(TEST_CLOOP_CELLSUM_ATTACK).

**What the cell-sum genuinely buys, and what it does not.** The cell-sum **does** kill the UV ($a\to0$)
divergence — the continuum-as-source-of-divergence assumption is genuinely false, and dropping it is banked
(ROOT-1). But finiteness does **not** imply uniqueness: a finite supertrace of a log-running 6-D determinant
still carries a **log-scheme degree of freedom** $\mu_{\rm cell}$. Three residual ambiguities survive:
(3a) the cutoff is a mode-**COUNT**, but $a_6$ needs a mode-**SCALE**; (3b) the η-dependence the ledger flagged
IS the scheme knob; (3c) supertrace ordering / sign is not fixed by a count. So:

$$ c_{\rm loop} = f(\text{frozen }K_6\text{ spectrum},\ \chi=-3,\ e^{-6\sigma}\text{ scaling}) \times [\text{one log-scheme factor fixed by }\mu_{\rm cell}]. $$

**Named axiom it could reduce to (κ³/π-clean).** This is exactly the GRANULARITY root's residue:

> **AXIOM-UNIFORM-CELL-VALUE** (target-blind form): *"there exists a single uniform operational cell
> $\Delta_0>0$ on the operational distinguishability metric (existence is the GRANULARITY root,
> `AXIOM-COSTFLOOR`; the spectral VALUE of $\Delta_0$ is a FLOOR-VALUES-RESIDUE on the same footing as
> $\hbar, k_B$), and $\mu_{\rm cell}$ is its spectral value at the $K_6$ scale."* Writable with no Λ and no
> hierarchy value. PASSES the kill-test as an *existence + value-residue* statement.

**The κ³/π pivot (the firewall, already run).** The single anchor that would pin $\mu_{\rm cell}$ is **one
measurement fixing $\Delta_0$ in spectral units at the $K_6$ scale** — and the *only* such anchor available is
the moduli-stabilization point $\partial_\sigma V=0$. **But $\partial_\sigma V=0$ IS the electroweak
hierarchy**, so anchoring $\mu_{\rm cell}$ there and then "predicting" the hierarchy is **tuning to the known answer —
explicitly FORBIDDEN** (the CLOSURE_CAMPAIGN firewall designated $\partial_\sigma V=0$ the forbidden anchor;
the earlier TEST_CLOOP suggestion to use it is superseded). The honest anchor is $\Delta_0$'s spectral value as
an *independent* floor-value input, on the footing of $\hbar$ — which the corpus does **not** supply.

**Specialist target / owner artifact.** (1) **Owner artifact:** the KK eigenvalue file / `reproduce_all.py`
(R0.7) that would pin $\mu_{\rm cell}$ *if* an independent floor-value for $\Delta_0$ existed. (2) **Specialist
target (the deep one):** "Compute the 13D FRG-2 β-vector and hence $c_{\rm loop}=\mathrm{tr}[a_6]$ of the
σ-fluctuation determinant on the frozen $K_6$ branch, target-blind." This is the shared Gap-04 / hierarchy /
a6 wall.

**Success ladder.**
- DERIVED-CLOSED: the FRG-2 β-vector computed → $c_{\rm loop}$ sign decided → the singlet Hessian eigenvalue
  fixed. **Blocked on the deep 13D FRG wall; not reachable here.**
- AXIOM-CLOSED: AXIOM-UNIFORM-CELL-VALUE named, with $\mu_{\rm cell}$ conceded as a floor-value residue (no new
  knob, no target). **Honest ceiling; already the GRANULARITY root's standing.**
- **sharper-OPEN / BLOCKED_INPUTS (current standing):** the cell-sum lowers the wall (finite, single-parameter)
  but does not remove it; the one residual $\mu_{\rm cell}$ has no $v$-independent readout.
- REFUTED: the FRG computation gives a definite $c_{\rm loop}$ with the wrong sign → the breathing singlet is a
  saddle → chamber-center is not a minimum.

**Honest disposition: BLOCKED_INPUTS, reducible to AXIOM-UNIFORM-CELL-VALUE.** The breathing-singlet Hessian is
the same FRG-2 wall as the hierarchy; the cell-sum converts "uncomputed continuum determinant" into "finite
cell-sum + one named floor-value anchor," which closes nothing but names the bottom precisely. **The
$\partial_\sigma V=0$ anchor is forbidden** (it is the hierarchy).

---

### 4.4 R6 — The Casimir net sign (the only route that could flip R3)

**This is the dependent of R3, broken out because it is the single decision that flips saddle→minimum.** Fully
covered as the rescue analysis under §4.2: per-sector sign LOCKED (bosons +, fermions −, scheme-independent);
**net sign = boson-surplus − fermion-surplus over the absent admissible-rep multiplicity table**; magnitude
threshold $\mu_{\rm cell}\cdot q\gtrsim0.5$.

**Technique: owner artifact (machine-lane) + T7.** Mount the multiplicity table + regularized zeta; form the
graded doublet-weighted sum; decide the net sign and magnitude.

**The κ³/π guard.** The frozen heat-kernel ledger weakly favors boson dominance (~3:1), but the opposing frozen
fact is $\chi=-3$. **Banking "boson-dominated rescue" without the table is NEEDS-TUNING** — the conjecture
$q>0$ is exactly the kind of plausible-but-unverified sign that the kill-test forbids until the table is
computed. The magnitude carries $\mu_{\rm cell}$ (shared with R1/R5), which has no $v$-independent readout.

**Success ladder.** **BLOCKED_INPUTS until the multiplicity table is mounted** → then **VERIFIED-RESCUE** (net
$>0$, $\mu_{\rm cell}\cdot q\gtrsim0.5$ → minimum) or **SADDLE-CONFIRMED** (net $\le0$ → F.9.7(2) fires).
**Highest value-per-effort *physics* item if the table exists** — it converts the named saddle attack surface
from "leaning saddle" into a decided verdict with no new tuning.

**Honest disposition: BLOCKED_INPUTS / NEEDS-TUNING.** The rescue is conditionally available but unbankable
without the absent multiplicity table; do not assert boson dominance as a result.

---

### 4.5 R2 — Global stabilization is an explicit NON-claim

**This is a claim-boundary residual, not a physics flip.** Global stabilization (no flat directions anywhere in
the active-branch moduli space) is **explicitly NOT claimed** (F.9.4 row 1 / F.10.1) — outside-chamber
configurations are *rejected by admissibility, not stabilized*. The task is not to "solve" it but to keep the
boundary honest and to state what *would* close it.

**Technique: T6 (all-operator-conditional) via the manuscript's own F.10.2 vocabulary rule.** The
required-vocabulary substitution rule already bans "full stabilization" / "no flat directions" / "all moduli
stabilized" in body text without the chamber-restriction qualifier. Compliance is mechanical.

**Named principle (κ³/π-clean):** **AXIOM-CHAMBER-RESTRICTION** — *"outside-chamber configurations are
rejected by the admissibility rulebook $\mathcal{C}_{\rm admiss}$ (an $\oplus$-layer rule), not dynamically
stabilized; the stabilization claim is scoped to the moduli used downstream by Gates 1–10."* This is the
F.9.4/F.10.3 statement; it carries no target value and is the honest scope wall.

**What would close it (and why it is SHARED-OPEN).** Global stabilization is the textbook hard problem on which
**all** rivals struggle (Battle verdict: TIE/SHARED-OPEN — *not* a Fable win). The downstream four-term
$V(\sigma)$ discharge (Paper IV) is the only object that even attempts the volume modulus globally, and it is
EXTERNAL with its own caveats and "promotions: zero" discipline; it changes **nothing** in this gate's
certificate (the F.8 downstream-corpus pointer says so explicitly).

**Success ladder.** AXIOM-CLOSED is not applicable as a *flip* (no axiom closes global stabilization here). The
endpoint is **DISCLOSED-CONSISTENT**: the non-claim is explicit, the vocabulary rule is enforced, and the
SHARED-OPEN status is stated honestly. **The cleanest, lowest-risk honesty item.**

**Honest disposition: DISCLOSED non-claim; keep it disclosed.** Do not let any downstream cross-reference
(Paper IV) be cited to upgrade this gate's certificate.

---

### 4.6 R4 — Witnesses fix moduli WITHIN the selected chamber (given-E)

**The obstruction.** The witnesses certify that the *selected* geometry is internally controlled; they do
**not** prove the witnesses *determine* the geometry. The chamber, the bundle $E$, and the family index
$\chi=-3$ are inputs (from SG-1/SG-3). Selection ≠ derivation; given-E ≠ derivation of E.

**Technique: T1 (axiom-floor), shared with SG-1/SG-3.** The honest reduction names the inheritance:

> **AXIOM-GIVEN-E-CHAMBER** (κ³/π-clean): *"the stabilization witnesses operate on the selected 13D K₆ branch
> and the supplied SM chiral content E; SG-6 certifies internal moduli-control of that selected geometry, not
> a cross-geometry determination."* Writable with no target value — it is a scope statement inherited from
> SG-1 (DECLARED-FROZEN, SHAPE selected-not-forced-absolute, ~9–10 injected reals beyond the 4 anchors) and
> SG-3 (DERIVED-GIVEN-E).

**Specialist target.** None unique to SG-6 — this is the SG-1 SHAPE-minimality / SG-3 given-E conditionality
flowing downstream. The contained SG-6 strengthening is R7 (the modular-fixed-point uniqueness), which is the
one witness that *could* be promoted from "selected" to "symmetry-forced."

**Success ladder.** AXIOM-CLOSED (AXIOM-GIVEN-E-CHAMBER named) is essentially already standing via the SG-1/SG-3
labels. DERIVED-CLOSED is not available at SG-6 (it would require closing SG-1's SHAPE-forcedness, which is
SELECTED-not-forced-absolute). **Honest disposition: AXIOM-CLOSED at the inherited scope axiom; the
within-chamber conditionality is the load-bearing caveat of the gate.**

---

### 4.7 R7 — $\tau=\omega$: strengthen "non-zero potential" to symmetry-uniqueness

**The obstruction.** F.2 asserts that perturbations off $\tau=\omega$ "produce a non-zero potential" restoring
$\tau\to\omega$, but the potential is **not computed**. The leg should rest on **modular-symmetry uniqueness**
of $\omega$ — a fixed point of an order-three modular symmetry is *symmetry-protected*, not tuned — which is
stronger than "a potential we did not compute."

**Technique: T1 (axiom-floor) — name the symmetry, prove the uniqueness.**

> **AXIOM-MODULAR-FIXED-POINT** (κ³/π-clean): *"the $F^+$ Cartan-torus modulus sits at the order-three modular
> fixed point $\tau=\omega=e^{2\pi i/3}$, the unique modular-symmetric / Weyl-rigid point of the chamber."*
> Writable with no flavor or hierarchy number — a symmetry statement. **Stronger than "read from a minimum"**
> because an order-three fixed point is symmetry-protected.

**Specialist target.** Hand the moduli specialist: "Show $\tau=\omega$ is the **unique** fixed point of the
$F^+$ Cartan-torus modular group, independent of any one-loop potential." A clean, bounded group-theory claim.

**Math to attempt.** Enumerate the fixed points of the chamber's modular group on the upper half-plane;
confirm $\omega$ is the order-three fixed point and characterize uniqueness within the chamber.

**Success ladder.** AXIOM-CLOSED (AXIOM-MODULAR-FIXED-POINT named; symmetry-protected) is the realistic
endpoint; DERIVED-CLOSED (uniqueness theorem proven) closes R7 outright and converts the $\tau$-witness from
"potential we did not compute" to "symmetry-forced point." sharper-OPEN if the fixed point is non-unique.
**Honest disposition: AXIOM-CLOSED likely, DERIVED if uniqueness is proven; this is the one witness that can be
genuinely strengthened by a contained computation.** (Note: this also strengthens SG-8's inherited
$\tau=\omega$ soft spot.)

---

### 4.8 R8 — No-tachyon check is conditional on the declared spectrum

**The obstruction.** F.4 asserts no tachyonic mode "on the active branch" at the declared witnesses, but a
deformation away from a witness could produce one (and then the certificate fails). The check is for the frozen
spectrum, not proven stable under the deformation classes the gate exposes it to.

**Technique: T6 (all-operator-conditional) within the admissibility-bounded deformation class.** The
deformations the gate must survive are bounded by admissibility (off-chamber ⇒ branch eliminated). The honest
statement is conditional, and the F.5 sensitivity analysis already bounds the in-chamber responses.

**Named principle (κ³/π-clean):** **AXIOM-NO-TACHYON-IN-CHAMBER** — *"on the active branch at the declared
witnesses, and under admissibility-bounded in-chamber deformations, no tachyonic mode appears."* Carries no
target value; it is the F.4 statement scoped to the admissible deformation class.

**Success ladder.** **DISCLOSED-CONSISTENT** (already substantially in place via F.4 + F.5). The residual is a
consistency sweep over the admissibility-bounded deformations, not a closure. (Note: R3's saddle finding is the
*spatial* (Hessian) analog of the tachyon check — a saddle direction is a flat/negative mass direction in field
space; R8 and R3 are linked, and a confirmed shape-doublet saddle would put pressure on R8 in that direction.)

**Honest disposition: DISCLOSED-CONSISTENT, conditional on the declared spectrum; dependent on R3.**

---

### 4.9 R9 — Phenomenological sufficiency inherits the downstream AUDIT (Gate-7 radii/spectrum)

**The obstruction.** The "residual moduli don't spoil Gates 1–10" claim relies on the downstream harnesses that
consume the SG-6 radii and threshold spectrum — chiefly the **Gate-7 threshold pipeline**, which is itself
AUDIT (the three δ are injected reals quoted to five sig figs but not regenerated; the G.3.2a row formulas do
not reproduce the decimals; `reproduce_all.py` + the ledger CSVs are ABSENT). SG-6's sufficiency leg is only as
machine-real as those harnesses.

**Technique: owner artifact (machine-lane), not an axiom.** A fail-closed reproducibility task, not a physics
closure.

**Owner artifact needed.** (1) Mount the Gate-7 ledger CSVs (`appendix_F_heat_kernel_ledger.csv`,
`appendix_F_threshold_outputs.csv`). (2) Run `reproduce_all.py` against the frozen radii + spectrum,
target-blind. (3) Confirm the threshold vector + unification residual regenerate. (4) Confirm a small
$\delta$ on the SG-6 radii moves the Gate-7 output within the F.3 $\delta$-response column.

**Math to attempt.** None new — execution + verification. Mechanical and fail-closed: if a perturbation of an
SG-6 radius moves a Gate 1–10 output beyond tolerance and is not bounded by F.3, phenomenological sufficiency
fails (F.10.4-3) and the offending downstream gate downgrades.

**Success ladder.** **BLOCKED_INPUTS until the Gate-7 CSVs/script are mounted** → then **VERIFIED** (sufficiency
machine-real) or **REFUTED** (a perturbation breaches tolerance → downgrade). **Value-per-effort high, but the
artifact is owned by SG-7** — closing SG-7's δ-harness AUDIT closes this too.

**Honest disposition: AUDIT, inherited from SG-7; closes when SG-7's reproduction harness is mounted.**

---

### 4.10 Attack-plan roll-up

| Residual | Technique | Named axiom (κ³/π-clean) | Specialist target / owner artifact | Realistic endpoint |
|---|---|---|---|---|
| R1 θ_H⋆ read from minimum | T5 + T7 | AXIOM-VEW-SECOND-ANCHOR | target-blind θ_H⋆ derivation; $v$-independence audit | AXIOM-CLOSED (2nd anchor, principled); derived-small OPEN/RELOCATION |
| R3 diagnostic Hessian (saddle surface) | T7 (symmetry split) | AXIOM-SHAPE-DOUBLET-CASIMIR | admissible-rep multiplicity table + zeta | sharper-OPEN, **saddle-leaning** ($-1$ curvature) |
| R5 $c_{\rm loop}$ / breathing singlet | T7 (cell-sum) | AXIOM-UNIFORM-CELL-VALUE | 13D FRG-2 β-vector; KK eigenvalue file | BLOCKED_INPUTS → AXIOM-CLOSED (floor-value residue) |
| R6 Casimir net sign (R3 flip) | owner artifact + T7 | (uses AXIOM-SHAPE-DOUBLET-CASIMIR) | multiplicity table | BLOCKED_INPUTS / NEEDS-TUNING |
| R2 global-stabilization non-claim | T6 (vocab rule) | AXIOM-CHAMBER-RESTRICTION | F.10.2 vocabulary sweep | DISCLOSED-CONSISTENT (SHARED-OPEN) |
| R4 given-E / within-chamber | T1 | AXIOM-GIVEN-E-CHAMBER | (inherited SG-1/SG-3) | AXIOM-CLOSED (scope axiom) |
| R7 $\tau=\omega$ strengthening | T1 | AXIOM-MODULAR-FIXED-POINT | modular-fixed-point uniqueness | AXIOM-CLOSED → DERIVED if unique |
| R8 no-tachyon conditional | T6 | AXIOM-NO-TACHYON-IN-CHAMBER | in-chamber deformation sweep | DISCLOSED-CONSISTENT (dep. R3) |
| R9 sufficiency AUDIT | machine-lane | (none) | mount Gate-7 CSVs + `reproduce_all.py` | BLOCKED_INPUTS → VERIFIED |

**REDUCE-vs-RELOCATE verdict on the plan.** The plan does **not** turn one hard problem into three harder ones.
Crucially, the two genuine-physics fronts collapse onto **one shared object**: $\mu_{\rm cell}$, the spectral
value of the uniform operational cell $\Delta_0$. R1 (the hierarchy via θ_H⋆), R3/R6 (the doublet Casimir
magnitude), and R5 (the breathing-singlet $c_{\rm loop}$) **all** reduce to "what is $\mu_{\rm cell}$, and can
it be read $v$-independently?" — and the already-run firewall answers: **no $v$-independent readout exists; the
only anchor ($\partial_\sigma V=0$) IS the hierarchy.** So the plan *reduces* the gate to a single named
floor-value residue plus a SADDLE-or-rescue verdict gated on one absent multiplicity table — it does not
multiply the difficulty. Each path either (a) names a single target-blind axiom that pays a debt in plain
sight (R1, R2, R4, R5, R7, R8), (b) is a bounded owner-artifact computation that decides a sign (R3/R6 the
multiplicity table; R9 the Gate-7 CSVs), and every $\mu_{\rm cell}$-touching path carries the explicit κ³/π
kill-test (the $\partial_\sigma V=0$ anchor is forbidden) so a circular closure cannot be banked. The honest
expected outcome of a full campaign: **~4 AXIOM-CLOSED (R1 as 2nd-anchor, R4, R5 to the cell residue, R7), ~2
DISCLOSED-CONSISTENT (R2, R8), 1 sharper-OPEN/saddle-leaning (R3), and 2 BLOCKED_INPUTS pending owner artifacts
(R6 multiplicity table, R9 Gate-7 CSVs).** No DERIVED-CLOSED is promised; the gate would move from
PARTIAL-with-diagnostic-Hessian to **PARTIAL-with-a-named-axiom-floor + a decided-or-saddle Hessian verdict** —
a real honesty/structural gain, **not** a promotion. **The single most likely *negative* — and the most
valuable one — is R3 firing: the shape-doublet saddle, if the Casimir rescue fails, REFUTES the
moduli-mass-generation row and downgrades it per F.10.4-2. That is the gate working as designed.**

---

## A. Anchoring & Hardening Map

This section runs SG-6 through **our internal honesty methodology** — the same hardening method that produced the live
**[Gaps & Walls Register](https://physics.magflowmeters.com/articles/GAPS_AND_WALLS_REGISTER.html)**: **classify each residual as a *gap* (a route
exists; what is missing is a finished computation or a measured input) or a *wall* (the route itself is the
problem) → hunt the implicit assumption it hides → name the measured-invariant truth it must terminate on → assign
the most conservative defensible disposition** — exactly that method (the framework, the gap-vs-wall split, and the
disposition vocabulary are defined in the linked Register). The discipline is honest bookkeeping, not closure: *no theorem breaks a wall*, and
the ceiling on every line below is **serious candidate / NOT validated**. The only existing measured anchors a
residual may terminate on are **{ℏ, M_Pl, spectrum-E, α_i(M_Z), y_t, |V_us|}**; anything else must name a NEW
invariant, hinge on a SCHEME-anchor, be a COMPUTATION-DEBT, or have NO witness → OPEN. Disposition words are used
exactly as the Register defines them: **DERIVED / DERIVED-GIVEN-E / AXIOM-CLOSED / DISSOLVED / SCHEME-ANCHORED /
OPEN / BLOCKED / measured-but-irreducible**.

### A.1 Per-residual anchoring map (one row per §3.1 residual)

| Residual (§3.1) | Gap or Wall (+kind) | Measured-invariant truth it must terminate on | Honest disposition | What would HARDEN it (concrete next step) |
|---|---|---|---|---|
| **R1 — θ_H⋆ ≈ 2.46×10⁻¹⁴ read from the V_Hos minimum** | **WALL** (SCALE / hierarchy wall; the only anchor *is* the answer) | **No existing anchor reaches it; needs a NEW named invariant.** Carries ~85% of the EW hierarchy via v_EW = θ_H⋆/(2πR_γ); v_EW is a *second dimensionful anchor* that cannot be built from {M_Pl, ℏ, dimensionless frozen geometry} without a σ-carrying length (Buckingham-π). It is **measured-but-irreducible** — v_EW=246 GeV is a datum, not a deduction here. | **AXIOM-CLOSED** at **AXIOM-VEW-SECOND-ANCHOR**; θ_H⋆ as "derived-small" stays **OPEN/RELOCATION** (Path A circular, Path B wrong-mechanism — firewall already run). | Hand a specialist the *target-blind θ_H⋆ derivation theorem*: compute θ_H⋆ from {γ, n_H=1, η_BK, R_γ} + a UV reference frozen **before** v_obs; if it provably needs ∂_σV=0 (the hierarchy), R1 is confirmed RELOCATION and the axiom stands. |
| **R2 — Global stabilization is an explicit NON-claim** | **WALL** (textbook SHARED-OPEN; rivals also struggle) | **Terminates on nothing measurable here** — it is a *claim-boundary* statement, not a number. No anchor; disclosure only. | **DISSOLVED-as-non-claim → DISCLOSED-CONSISTENT.** Not AXIOM-CLOSED as a flip; no axiom closes global stabilization. Outside-chamber configs are rejected by admissibility, *not stabilized*. | Mechanical: enforce the F.10.2 vocabulary rule (ban "full stabilization"/"no flat directions" without the chamber qualifier) and keep any Paper-IV V(σ) cross-reference from being cited to upgrade this certificate (F.8). |
| **R3 — Positive-definite Hessian is DIAGNOSTIC-ONLY (named saddle attack surface)** | **GAP** (computation-debt; one Λ-free piece already done) | **SCHEME-anchored + COMPUTATION-DEBT.** The Λ-free shape-doublet curvature is already **−1 (SADDLE)** target-blind; the breathing singlet rides the uncomputed c_loop. No existing measured anchor; the rescue magnitude carries μ_cell (a scheme object). | **OPEN — saddle-leaning** (sharper-OPEN). The named F.9.7(2) falsifier is the *closest to firing* of any SG-6 residual. | Mount the **admissible-rep multiplicity table** (ABSENT) + regularized zeta; compute sgn(boson−fermion surplus) and the magnitude μ_cell·q≳0.5 → decide rescue vs SADDLE-CONFIRMED. **Do NOT bank boson-dominance without the table.** |
| **R4 — Witnesses fix moduli WITHIN the selected chamber (given-E)** | **WALL** (inherited SHAPE/given-E wall from SG-1/SG-3) | **Terminates on spectrum-E** — the SM chiral content + χ=−3 are supplied as input; witnesses operate *on* E, they do not derive it. given-E ≠ derivation of E. | **AXIOM-CLOSED** at **AXIOM-GIVEN-E-CHAMBER** (already standing via SG-1 DECLARED-FROZEN / SG-3 DERIVED-GIVEN-E). This is the load-bearing conditionality of the whole gate. | Nothing closes it at SG-6 (it is SG-1 SHAPE-forcedness, SELECTED-not-forced-absolute). The only contained SG-6 strengthening is R7 (promote one witness from "selected" to "symmetry-forced"). |
| **R5 — c_loop / breathing-mode singlet = the 13D FRG-2 β-vector wall** | **WALL** (GRANULARITY-root residue; shared deep FRG-2 wall w/ Gap-04, a6) | **Terminates on a NEW invariant** — the spectral VALUE of the uniform operational cell Δ₀ (μ_cell), a floor-value residue *on the footing of ℏ* but **not supplied by the corpus**. The only available anchor (∂_σV=0) IS the hierarchy → **forbidden** (circular). | **BLOCKED_INPUTS**, reducible to **AXIOM-CLOSED** at **AXIOM-UNIFORM-CELL-VALUE**. The cell-sum kills the continuum divergence (DISSOLVED leg) but leaves one log-scheme factor μ_cell → **SCHEME-ANCHORED** residue. | The deep specialist target: compute the **13D FRG-2 β-vector** → c_loop = tr[a₆] of the σ-fluctuation determinant, target-blind. Short of that, supply an *independent* floor-value for Δ₀ (footing of ℏ) — **never** anchor μ_cell at ∂_σV=0. |
| **R6 — Casimir net sign (the only route that could flip R3)** | **GAP** (BLOCKED-INPUTS; one machine-lane decision) | **SCHEME-anchored** (per-sector sign LOCKED by geometry; net sign = graded boson−fermion surplus over the **ABSENT** multiplicity table; magnitude carries μ_cell, no v-independent readout). | **BLOCKED_INPUTS / NEEDS-TUNING.** The rescue is conditionally available but **unbankable** — banking boson-dominance (~3:1 heat-kernel) without the table is the κ³/π-forbidden plausible-sign. χ=−3 is the opposing frozen fact. | Mount the multiplicity table → form Σ(−1)^F d·n·(w₁²+w₂²) → VERIFIED-RESCUE (net>0, μ_cell·q≳0.5) or SADDLE-CONFIRMED (net≤0 → F.9.7(2) fires). Highest physics value-per-effort *if* the table exists. |
| **R7 — τ=ω "non-zero potential" leans on an uncomputed potential** | **GAP** (bounded group-theory; cleanest witness) | **Terminates on a symmetry invariant** — order-three modular fixed-point uniqueness, *not* a measured number. Symmetry-protected, not read from a curve. | **AXIOM-CLOSED** at **AXIOM-MODULAR-FIXED-POINT**, → **DERIVED** if uniqueness is proven. The one witness genuinely upgradable by a contained computation. | Prove τ=ω is the **unique** fixed point of the F⁺ Cartan-torus modular group, independent of any one-loop potential — converts the leg from "potential we did not compute" to "symmetry-forced point." (Also hardens SG-8's inherited τ=ω soft spot.) |
| **R8 — No-tachyon check is conditional on the declared spectrum** | **GAP** (consistency sweep; admissibility-bounded) | **No new anchor** — conditional on the frozen spectrum + admissibility-bounded in-chamber deformations; F.5 already bounds in-chamber responses. | **DISCLOSED-CONSISTENT** at **AXIOM-NO-TACHYON-IN-CHAMBER**; dependent on R3 (a confirmed shape-doublet saddle is the field-space analog of a tachyon and would put pressure here). | Run the consistency sweep over the admissibility-bounded deformation class; couple the verdict to R3's Hessian-sign outcome (saddle ⇒ pressure on the no-tachyon leg in that direction). |
| **R9 — Phenomenological sufficiency inherits downstream AUDIT (Gate-7 radii/spectrum)** | **GAP** (COMPUTATION-DEBT; owned by SG-7) | **Hinges on a SCHEME-anchor inherited from SG-7** — Gate-7 is now **Diagnostic-only**, its threshold vector (+4.8424, −3.1112, −1.7313) **fitted-to-target, not formula-derived**; signs geometric, magnitudes SCHEME-ANCHORED. SG-6's sufficiency is only as machine-real as that harness. | **AUDIT**, inherited from SG-7 (**SCHEME-ANCHORED / BLOCKED_INPUTS**). Closes when SG-7's reproduction harness is mounted; cannot be more real than its upstream. | Mount the Gate-7 ledger CSVs + `reproduce_all.py` (ABSENT); regenerate the threshold vector + unification residual target-blind; confirm a small δ on the SG-6 radii moves Gate-7 output within the F.3 δ-response column. (Owned by SG-7 — closing SG-7's δ-harness AUDIT closes this.) |

### A.2 Gate-level rollup

- **Overall disposition.** SG-6 is **PARTIAL → conservatively OPEN** (matching the live Register's SG-6 line). The
  genuine, strong content is the **witness-type predicate** (a *type* test that makes "a potential we minimized at
  the answer" fail by construction) plus the discrete/topological witnesses and the S₃-critical-point structure —
  these are real, target-blind facts, not promotions. The open surface terminates as **3 AXIOM-CLOSED** (R1 as a
  second irreducible anchor, R4 the given-E scope, R7 the modular fixed point — likely DERIVED), **1 reducible-to-
  AXIOM-CLOSED but BLOCKED** (R5 → AXIOM-UNIFORM-CELL-VALUE, gated on the FRG-2 wall), **2 DISCLOSED-CONSISTENT**
  (R2 global-stabilization non-claim, R8 conditional no-tachyon), **1 OPEN/saddle-leaning** (R3 — the named
  F.9.7(2) falsifier, *already* −1 on Λ-free geometry), and **2 BLOCKED_INPUTS pending owner artifacts** (R6 the
  multiplicity table, R9 the SG-7 reproduction harness). **No DERIVED-CLOSED is promised; No status was ever upgraded.**
- **Anchors it depends on.** **spectrum-E** (R4: the gate certifies internal moduli-control *given* the supplied
  chiral content + χ=−3, never derives it) and **ℏ-footing only** (R5: μ_cell is a floor-value residue *on the
  footing of* ℏ, but its spectral value is **not supplied** — it is a NEW invariant, not the measured ℏ itself).
  It depends on **no** clean termination on M_Pl / α_i(M_Z) / y_t / |V_us|; the one residual that touches a
  *number* (R1, the hierarchy) terminates on **measured-but-irreducible v_EW**, not on an existing anchor. R9
  inherits SG-7's **SCHEME-anchor**, not a measured one.
- **Single highest-leverage hardening move.** **Mount the admissible-rep multiplicity table** (currently ABSENT).
  It is the one owner artifact that decides the net Casimir sign and thereby flips the named saddle attack surface
  (R3/R6) from "leaning saddle" to a *decided verdict* — VERIFIED-RESCUE (minimum) or SADDLE-CONFIRMED (F.9.7(2)
  fires, the moduli-mass-generation row downgrades per F.10.4-2). It is the highest physics value-per-effort step
  and the only one that converts a diagnostic into a falsifiable result **without new tuning**. The deeper twin
  wall (R1/R5, the hierarchy and the breathing singlet) collapses onto a **single object — μ_cell** — which the
  already-run firewall confirms has **no v-independent readout** (its only anchor ∂_σV=0 IS the hierarchy; the
  κ³/π kill-test fires), so that front hardens only to a *named axiom floor*, never to a derivation.

### 🎯 Target anchor(s) for this gate

The body of SG-6 carries **no new measured anchor** and terminates on nothing of its own: the chamber witnesses
fix the moduli **WITHIN the already-selected chamber, given-E** (witness-*type* test, **not** a tuned potential
minimized at the answer) — so the witness ledger terminates on **spectrum-E** (R4, **AXIOM-CLOSED** at
AXIOM-GIVEN-E-CHAMBER) and on the **ℏ-footing only** (R5: μ_cell is a NEW invariant on the footing of ℏ, not the
measured ℏ — **SCHEME-ANCHORED / BLOCKED**), and **global stabilization is an explicit NON-claim** (R2,
DISCLOSED-CONSISTENT — it terminates on no number). The **one acute open leg is θ_H⋆ ≈ 2.46×10⁻¹⁴**, *read from the
chamber one-loop V_Hos minimum* and carrying **~85% of the EW hierarchy** via v_EW = θ_H⋆/(2πR_γ): its target anchor
is **v_EW = 246 GeV**, which here is **measured-but-irreducible** (a second dimensionful anchor, not deducible from
{M_Pl, ℏ, dimensionless frozen geometry}) — θ_H⋆ as "derived-small" stays **OPEN/RELOCATION**, AXIOM-CLOSED at
AXIOM-VEW-SECOND-ANCHOR. **No new anchor is introduced; No status was ever upgraded.** The hardening move is **not** a new datum
but the **θ_H⋆ / Hessian derivation** — a target-blind θ_H⋆ from {γ, n_H=1, η_BK, R_γ} + a pre-v_obs UV reference,
plus the positive-definite Hessian decision at (1,1,1) gated on the ABSENT admissible-rep multiplicity table (R3/R6).

---

## 5. References & source map

### 5.1 Website source-of-truth (common material — link, don't duplicate)

- **Paper I, GUT.html** — <https://physics.magflowmeters.com/articles/GUT.html>
  - **§6.6** Gate-6 card (binding status: *Claimed certificate pass under declared admissibility and
    moduli-control assumptions*); **§6.12** falsification map (Gate 6 row: *Diagnostic only* if Hessian
    falsified / *Open* if phenomenological sufficiency falsified); **§6.13** certificate summary.
  - **§5.5** narrative module; the **$y=R^2$ curve-not-a-number** discipline (§6 opening of the gate index).
  - **Appendix F** (Stabilization — the binding authority for Gate 6): **F.1** stabilization claim; **F.2**
    potential/witness structure; **F.3** the 12-row moduli ledger; **F.4** spectrum + tachyon check; **F.5**
    sensitivity; **F.6** certificate JSON; **F.9** Stabilization-vs-Admissibility (F.9.1 five concepts, F.9.2
    what is proved, **F.9.4 the NON-claim**, **F.9.5 the Hessian diagnostic**, F.9.6 required vocabulary,
    **F.9.7 falsification paths**); **F.10** claim-type ledger (F.10.1 ledger, **F.10.2 vocabulary substitution
    rule**, F.10.3 binding statement, F.10.4 reviewer falsification path).
  - **Appendix CR module CR6** — reader companion (explanatory only; mints no status).
  - **§6.8 / Appendix H** — the Gate-8 Hosotani object on which R1 (θ_H⋆ / $v_{\rm EW}$) rides; **Appendix G**
    — the Gate-7 threshold pipeline on which R9 (sufficiency) depends.
- **Paper IV, TOE.html** — <https://physics.magflowmeters.com/articles/TOE.html> (the four-term $V(\sigma)$
  discharge; EXTERNAL; cross-reference only — changes nothing in this gate's certificate per F.8).

This dossier recaps only what is needed to attack; the site controls all common material.

### 5.2 Corpus locations (authoritative inputs to this dossier)

| Source | Path | Role |
|---|---|---|
| Per-gate dossier spec | `…/rendered/TOE/PER_GATE_DOSSIER_SPEC.md` | structure (sections 0–5) |
| SG-6 status line | `…/rendered/TOE/SCOPED_GUT_PER_GATE_STATUS_LEDGER.md` | the PARTIAL label + honest caveats (carried verbatim) |
| **Λ-anchored moduli test** | `…/rendered/TOE/TEST1_LAMBDA_ANCHORED_MODULI_STABILIZATION_2026-06-23.md` | Step A/B/C verdict; chamber V(u) is fiat-not-minimized; the four-term V(σ) is σ-only; counting/under-determination wall |
| **Λ-free chamber potential (S₃ split)** | `…/rendered/TOE/TEST2_LAMBDA_FREE_CHAMBER_POTENTIAL_2026-06-23.md` | the $3=\mathbf1\oplus\mathbf2$ decomposition; $(1,1,1)$ = $S_3$-critical-point free; singlet=wall, doublet=computable |
| **Shape-doublet Hessian** | `…/rendered/TOE/TEST_SHAPE_DOUBLET_STABILITY_2026-06-23.md` | the Λ-free doublet curvature Hessian $=-1$ (SADDLE); structure-constant triangle is the destabilizer |
| **Casimir rescue test** | `…/rendered/TOE/TEST_CASIMIR_RESCUE_SHAPE_DOUBLET_2026-06-23.md` | per-sector sign LOCKED; net sign needs the ABSENT multiplicity table; $\mu_{\rm cell}\cdot q\gtrsim0.5$ |
| **Cell-sum c_loop attack** | `…/rendered/TOE/TEST_CLOOP_CELLSUM_ATTACK_2026-06-23.md` | c_loop = finite cell-sum + one log-scheme factor $\mu_{\rm cell}$; uniform-$\Delta_0$ residue |
| **SCALE/hierarchy final verdict** | `…/rendered/TOE/SCALE_HIERARCHY_FINAL_VERDICT_2026-06-23.md` | θ_H⋆ ≈ 2.46e-14 carries 85% hierarchy; Path A circular (Buckingham-π), Path B wrong-mechanism; $v_{\rm EW}$ a 2nd anchor; $\partial_\sigma V=0$ forbidden |
| Closure campaign (R1) | `…/rendered/TOE/CLOSURE_CAMPAIGN_RESULT_2026-06-24.md` | the decisive SCALE-firewall verdict (μ_cell no $v$-independent readout); κ³/π kill-test; AXIOM_CLOSED ≠ proven |
| Closure campaign (R2) | `…/rendered/TOE/CLOSURE_CAMPAIGN_ROUND2_2026-06-24.md` | demotion-on-verify norm; relabel-fail kill-test discipline |
| Axiom ledger | `…/rendered/TOE/AXIOM_LEDGER.md` | ROOT-1 `AXIOM-COSTFLOOR` (existence not value) + FLOOR-VALUES-RESIDUE ($\Delta_0$ value); ROOT-2 SHAPE (selection ≠ derivation) |
| GUT manuscript | `…/rendered/GUT/GUT.md` | §5.5 / §6.6 (Gate-6 card); Appendix F (F.1–F.10); R1.6 witness predicate; F.3 12-row ledger |

### 5.3 Frozen-object hash index (load-bearing; READ-ONLY)

Branch `dcc66f1b2685` · manifest meta `a5b1e6f9d951` · three Cartan radii `634438ce0776` / `2381d472c62e` /
`0e8b8dba2cf0` · chamber-center $\vec u=(1,1,1)$ (A1.2) · $\tau=\omega$ `03b30a9c931a` · $\eta_{BK}$
`84e94518d3f5` · RG transport `f531205a9159` · $M_Z$ `a6852c7a6b00` · $N_{u,d,e}$ `20dc4e0b8220` · $\theta_F$
`1ff57f48d45a` · spin-ℂ index / $\chi=-3$ (R1.4) `0fd19c9ae0c1` · threshold vector
$(\delta_1,\delta_2,\delta_3)=(+4.8424,-3.1112,-1.7313)$. **θ_H⋆ (Hosotani minimum), $\mu_{\rm cell}$ /
$\Delta_0$ spectral value, $c_{\rm loop}$, the admissible-rep multiplicity table, and $M_R$: no SG-6 hash —
read / uncomputed / ABSENT.**

---

### Closing honest statement

SG-6 is **PARTIAL**. Its genuine, defensible content is real and strong: the **witness-type predicate** — every
downstream-used modulus carries a structural witness of declared TYPE (Weyl-rigid chamber / modular fixed point
/ integer winding / discrete topology / boundary-condition pin), with "a potential we minimized at the answer"
failing **by construction** — backed by the **12-row witness ledger**, a **no-tachyon check**, a
**coverage lint**, and the honest **claim-type ledger** that fences exactly what is proved. Its open surface is
equally clear: **global stabilization is an explicit NON-claim** (F.9.4/F.10.1); the **positive-definite
Hessian is DIAGNOSTIC-ONLY** (F.9.5) — and on current Λ-free perturbative geometry the **shape-doublet Hessian
is a SADDLE** ($-1$, the named F.9.7(2) attack surface leaning toward firing, rescuable only by a Casimir term
whose net sign needs an ABSENT multiplicity table); the acute soft spot is **θ_H⋆ ≈ 2.46×10⁻¹⁴**, carrying
**~85% of the EW hierarchy**, **read from a one-loop minimum, not derived-small**; and the witnesses fix the
moduli **within the selected chamber** (given-E). The attack plan reduces these to named, target-blind axioms
and bounded owner-artifact computations — and crucially collapses the two genuine-physics fronts (the hierarchy
and the Hessian sign) onto **one shared object $\mu_{\rm cell}$**, which the already-run firewall confirms has
**no $v$-independent readout** ($\partial_\sigma V=0$, its only anchor, **IS** the hierarchy — the forbidden,
circular anchor; the κ³/π kill-test fires). The realistic ceiling is a named axiom floor (AXIOM-VEW-SECOND-ANCHOR,
AXIOM-UNIFORM-CELL-VALUE, AXIOM-MODULAR-FIXED-POINT) plus a **decided-or-saddle Hessian verdict** — **not** a
promotion; and the most valuable likely outcome is a *negative*: the shape-doublet saddle firing F.9.7(2) if
the Casimir rescue fails. **No status was ever upgraded; frozen branch `dcc66f1b2685` / `a5b1e6f9d951` READ-ONLY; given-E ≠
derivation of E; selection ≠ derivation; dissolved ≠ solved; nothing applied, nothing deployed.**

*Dossier built 2026-06-24. Our geometry (13D K₆ branch) only. Common material referenced to the published
website source-of-truth, not duplicated.*