SG-3 — Three Generations as a Topological Index $\chi(K_6,\mathcal{E})=-3$: Per-Gate Closure-Attack Dossier

What this is. The closure-attack packet for SG-3 (three chiral generations as a rigid topological index) on the frozen 13D $K_6$ branch ONLY. It recaps how our geometry closes the family count (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 index $\chi(K_6,\mathcal{E})=-3$ is a rigid deformation-proof integer with no dial (Atiyah–Singer), but it is computed WITH $\mathcal{E}$ (the SM chiral content) as input — it certifies this geometry+bundle yields three families, not that three is unique across all geometries; the count is read off a chosen bundle/weight, and the SHAPE finding is that families are bundle-selected on every carrier (the cheaper $\mathbb{CP}^2$ carrier produces "three by dial"; $K_6$'s forcedness is bought by paying +2 dimensions with a 3-generation–aligned tie-break — the suite's live smuggling surface); and a sharp technical finding the corpus already records: the literal "spin-$\mathbb{C}$" wording is REFUTED for pure-SM content (spin-$\mathbb{C}$ is provably obstructed; Davighi–Gripaios–Lohitsiri) — the genuine forced object is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$ bundle, which is FORCED-GIVEN-E and does not force $\mathcal{E}$. Closure paths must be non-target-fitted (the $\kappa^3/\pi$ falsification test: a proposed axiom counts only if it would be written WITHOUT knowing the answer). given-E $\neq$ derivation of E. AXIOM-CLOSED $\neq$ proven; selection $\neq$ derivation; dissolved $\neq$ solved.

0. Header & Verdict

0.1 Abstract — established / open / what would close it

Established (given-$\mathcal{E}$, given the selected & frozen geometry+bundle). On the frozen branch the family count is the integer index read by two complementary counting theorems (Appendix E.1; GP.3): the Borel–Weil–Bott (BWB) index of the frozen line/spinor bundle on the flag manifold $K_6=SU(3)/T^2$ returns $\chi(K_6,\mathcal{E})=-3$, and the Atiyah–Patodi–Singer (APS) boundary index on the orbifold $S_Y^{\,1}/\mathbb{Z}_2$ returns $(n_L,n_R)=(+3,0)$. The magnitude $\lvert\mathrm{Index}\rvert=3$ is the number of generations; the sign $-3$ and the one-sidedness $n_R=0$ are chirality (left-handed families, no surviving mirror). This is a rigid, deformation-proof integer: no continuous modulus can move it inside the declared search category — the genuine, defensible content. It is the kind of object a family count must be (a winding number, not a volume knob). The gate creates the surviving three-family multiplet list consumed downstream by Gates 5/7/9/10.

Open (the honest deductions). (1) given-$\mathcal{E}$ — the index is computed with the SM chiral content $\mathcal{E}$ (the bundle whose Chern class returns $-3$) supplied as input; it certifies this geometry+bundle yields three, not that three is unique across geometries, and emphatically not that $\mathcal{E}$ itself is forced (the corpus T3 — "E is forced by anomaly-freedom + minimality" — is REFUTED: anomaly-freedom is a filter, not a determiner; infinitely many anomaly-free chiral $U(1)$ extensions exist; the generation number is unfixed; and $\chi(K_6,\mathcal{E})=-3$ is circular for forcing $\mathcal{E}$ because $\mathcal{E}$ is its input). (2) the count is bundle-selected, not carrier-forced — "3" is read off a chosen first Chern class / weight; the cheaper $\mathbb{CP}^2=SU(3)/U(2)$ carrier produces the count as a continuous bundle-moduli choice ("three by dial"), and $K_6$'s forcedness is purchased by paying +2 dimensions under a tie-break (anti-fitting / "adjustable counts as fail") that is itself aligned with the 3-generation target — the SHAPE suite's explicitly named live smuggling surface. (3) bundle admissibility — which spin-$\mathbb{C}$/line-bundle data are admitted is set by no-fourth-generation + the hypercharge ledger + the $\mathbb{Z}_6$ centre; the index-changing deformation that would give 2 or 4 is excluded by the LEP $N_\nu$ bound, i.e. by data, not by an internal forcing of the bundle. (4) the literal "spin-$\mathbb{C}$" wording is REFUTED for pure SM (Davighi–Gripaios–Lohitsiri arXiv:1910.11277): the genuine global object forced by $\mathcal{E}$'s charges is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$ ($\mathbb{Z}_6$ via the Tong congruence $q\equiv 3z_2-2z_3\bmod 6$; $(-1)^F$ = the $SU(2)$ $2\pi$ rotation in the centre), FORCED-GIVEN-E, not spin-$\mathbb{C}$.

What would close it (the ladder). DERIVED-CLOSED would require a target-blind theorem that forces the family count to be exactly 3 across the declared carrier-space without the bundle as input — e.g. a bundle-uniqueness / "$\mathcal{E}$ forces the $K_6$ weight" theorem (the SHAPE color-rung bundle-admissibility target), or a proof that the carrier+admissibility data alone (no LEP input, no 3-aligned tie-break) return $\lvert\mathrm{Index}\rvert=3$ uniquely. The realistic ceiling per residual is AXIOM-CLOSED: name one explicit, target-blind posit (e.g. "the admissible bundle is the minimal-weight spin-$\mathbb{C}$/twisted-spin lift compatible with the $\mathbb{Z}_6$ centre"; "the carrier is the unique purely-abelian-isotropy clean $SU(3)$ flag") that pays the debt in plain sight. The most likely honest outcome on residuals (b)/(c) is sharper-OPEN: the family count stays a rigid index GIVEN a bundle that is selected, not forced.

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

The full construction, the worked index, the parity table, and the freeze records are the published manuscript, Paper I:

• GUT.html §6.4 (Gate-4 card — chirality / no mirrors / family count; status verbatim) — https://physics.magflowmeters.com/articles/GUT.html
• §5.3 narrative module (Gate 4); §3.4 (topological-count machinery); the Strongest-Objections block (Objection 4: "the three-family count is asserted by choosing the geometry"); §4.6 / §4.9 (the selector dimension-payment and the anti-fitting firewall); Appendix E (Chirality Closure — authoritative: E.1 mechanism, E.2 three generations, E.3 mirror ledger, E.6 certificate); Appendix CR module CR4 (reader companion); Appendices GP.3 (the counting theorems — Atiyah–Singer / APS / Borel–Weil–Bott), B1/B2 (selector / category-relativity), C2 ($K_6$ dossier), C4 ($S_Y^{\,1}/\mathbb{Z}_2$ dossier), A2.2 (bundle ledger); Appendix R0 freeze records. Same public article.
• Downstream cross-references (do not change this gate): TOE.html — https://physics.magflowmeters.com/articles/TOE.html.

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

1. How OUR geometry closes the family count — the closure claim (concise)

Full derivation: GUT.html §6.4 + §5.3 + Appendix E + GP.3 + CR4. This section is the attack-grade recap, not the derivation.

1.1 The mechanism end-to-end

The family count is generated by two index computations on the frozen geometry, one per half of the gate ("$K_6$ counts the families; $S_Y^{\,1}/\mathbb{Z}_2$ makes them chiral by projecting out the mirror copy" — CR4.6). The pipeline is one forward chain (§5.3, Appendix E.1):

 K6 = SU(3)/T2  +  frozen spin-C/twisted-spin bundle  --BWB index-->  chi(K6,E) = -3   --> |Index| = 3 families
 S1_Y/Z2 fold   +  frozen boundary parity ledger        --APS index-->  (n_L, n_R) = (+3, 0) --> mirrors removed
 E_matter (chiral actors) + P_chi projector             --zero modes--> three left-handed SM generations

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

1. The carrier $K_6=SU(3)/T^2$ (R1.4; dossier C2; hash of the bundle data 0fd19c9ae0c1). The complete flag manifold of $\mathbb{C}^3$: $\dim=8-2=6$, a compact homogeneous Kähler manifold with isometry group $SU(3)$ and Euler characteristic $\chi(K_6)=\lvert W(SU(3))\rvert=6$. It is the same factor Gate 2 used for color recovery — the strongest reuse point in the manuscript (color carrier $\to$ family-count carrier, no new term). Line bundles on $K_6$ are classified by weights; their cohomology — hence the particle content they induce — is computed in closed form by Borel–Weil–Bott (GP.3). (Attack handle: the carrier is forced as the unique clean abelian-isotropy $SU(3)$ flag, but the bundle/weight read on it is selected — residuals (b)/(c).)

2. The spin-$\mathbb{C}$ / twisted-spin bundle and its BWB index. Kähler manifolds such as $K_6$ are canonically spin-$\mathbb{C}$; on a homogeneous Kähler space the relevant index is computable in closed form, and the frozen bundle returns $\chi(K_6,\mathcal{E})=-3$, so $\lvert\mathrm{Index}\rvert=3$ (Appendix E.1; full bundle data A2.2; GP.3 Borel–Weil–Bott). The sign records chirality; the magnitude records the family count. No additional family appears (the index returns no further zero mode under the same bundle); no fewer (the integer is robust against continuous deformation within the declared category). (Attack handle: the word "spin-$\mathbb{C}$" is technically refuted for pure SM — the genuine object is twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$; residual R4. The value $-3$ is robust; the labelling of the structure is the soft spot.)

3. The no-mirror fold $S_Y^{\,1}/\mathbb{Z}_2$ and its APS index. A closed factor is handedness-neutral (fact F2: left/right come in matched pairs, net handed count zero — a bare $S_Y^{\,1}$ mirrors every fermion, APS index $(+3,+3)$). The repair is an edge: the orbifold $\theta\mapsto-\theta$ with fixed points $\{0,\pi\}$ (freeze R1.3 ac4d2df3e708) admits one handedness and refuses its mirror. Under the frozen parities the APS boundary index returns $(n_L,n_R)=(+3,0)$ — a one-sided count impossible on any closed factor. (Attack handle: a no-go leg; its force is structural — F2 plus the fold — and it is the cleaner half of the gate.)

4. The chiral actors $\mathcal{E}_{\rm matter}$ and the projector $P_\chi=\tfrac12(1+\gamma_5\Gamma_8)$ (Appendix C7; $P_\chi$ derived from ac4d2df3e708+0fd19c9ae0c1). The matter bundle is the $\otimes$-layer object on which $P_\chi$ and the index act; without it neither has a domain. The surviving content is exactly the SM set: the right-handed singlets ($u_R,d_R,e_R$, and the neutrino-sector mode) arise from the conjugate sector through the chamber projectors with the same generation count.

5. The generation module $\mathcal{G}_{\rm gen}=\mathrm{span}\{g_1,g_2,g_3\}$, $\dim=3$, matched to the family index $-3$ (A1.13; consumed by the $F^+$ flavor chamber, SG-8). SG-8 inherits this dimension; SG-3 is the gate that supplies "3" to flavor.

1.2 What "closed" means here, and its conditionality

"Closed" for SG-3 is a rigid integer family-count mechanism under declared assumptions — strictly not a from-nothing inevitability of three, not a proof that three is unique across all geometries, not a derivation of the SM chiral content $\mathcal{E}$. It is conditional on:

• given-$\mathcal{E}$ — the SM chiral content (the bundle $\mathcal{E}$ whose Chern class returns $-3$, and the hypercharge ledger that fixes admissibility) is supplied as input; SG-3 does not derive $\mathcal{E}$. (The index is the output of a computation whose input is $\mathcal{E}$ — using it to "force" $\mathcal{E}$ would be circular; the corpus says so explicitly, REDUCTION_SPINTWIST_C §3(iii).)
• given-the-selected-geometry+bundle — the frozen carrier $K_6$ and the selected bundle/weight; the carrier is forced (abelian-isotropy uniqueness; CP² over-produces gauge at Gate 2), but the bundle is selected inside a declared admissibility class, not proven unique.
• given-the-data-exclusion — the deformation that would change the index to 2 or 4 is excluded by the LEP $N_\nu$ bound ($2.984\pm0.008$ light species). The count's robustness against fourth-generation rescue is partly a data input, not a pure internal forcing.

The genuine, defensible content (the part that survives the honest accounting):

The conditional / non-genuine content (selection in disguise, or data-fed):

So the precise statement of closure: the geometry genuinely makes the family count a rigid deformation-proof integer of the correct kind (3, left-handed, no mirror) — but it does so by reading a bundle whose data is $\mathcal{E}$ itself; it certifies this geometry+bundle, not that three is forced across geometries, and not that the bundle is unique. given-E is the load-bearing qualifier of the whole gate.

2. Verify the status — is DERIVED-GIVEN-E 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

2.2 What reproduces, plainly

• The integer reproduces, and its rigidity is real. $\lvert\chi(K_6,\mathcal{E})\rvert=3$ is a Borel–Weil–Bott output — a closed-form rep-theory computation on $SU(3)/T^2$, not a numerical fit. The APS $(+3,0)$ is hand-checkable via the one-line interval toy. The bare-$S_Y^{\,1}$ control $(+3,+3)$ confirms the fold is load-bearing. This is the strongest, most defensible leg of the gate — and it is genuinely stronger than SG-8's strongest leg, because an integer index is more rigid than a $\kappa$-power ratio.
• The rep-theory engine reproduces target-blind. The $m_0(p,q)$ multiplicity rule and the $\mathbb{Z}_6$ Tong congruence both regenerate from scratch (W4/W5), computed from the actual SM charges, not reverse-engineered.
• The deformation-proof claim reproduces in-category. No continuous modulus moves the index inside the declared search category — that is what "topological" means here (CR4.4).

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

• The index does NOT reproduce without the bundle as input. $\chi(K_6,\mathcal{E})=-3$ is a function of $\mathcal{E}$ (the first Chern class / weight). Feed a different admissible weight and the index changes. So "three families" reproduces given-$\mathcal{E}$, and the gate's certificate is honest exactly to that conditioning — it is not a from-nothing reproduction of "3."
• The literal "spin-$\mathbb{C}$" structure does NOT reproduce — it is refuted. W6 shows pure-SM content provably forbids a spin-$\mathbb{C}$ structure; the genuine forced object is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$ (twist order $n=2$, $(-1)^F=$ the $SU(2)$ $2\pi$ rotation in the centre; REDUCTION_SPINTWIST_C §3). The value $\lvert\chi\rvert=3$ survives this correction (it is a count, not a structure label), but the manuscript's wording "spin-$\mathbb{C}$ index" is a structural overstatement that the corpus itself flags as a smuggle if asserted literally.
• The "three is forced" reading does NOT reproduce across geometries. The cross-geometry uniqueness of 3 is OPEN; the $\mathbb{CP}^2$ "three by dial" elimination depends on a 3-aligned tie-break (W7); the bundle is selected, not proven unique (residuals b, c).
• The "$\mathcal{E}$ is forced" reading is REFUTED. Corpus T3 ("E forced by anomaly-freedom + minimality") is explicitly refuted (REDUCTION_SPINTWIST_C §3(iii)): anomaly-freedom is a filter, the hypercharge family is 1-parameter, there are infinitely many anomaly-free chiral $U(1)$ extensions (Allanach et al.), the generation number is unfixed, and $\chi=-3$ is circular for forcing $\mathcal{E}$. E remains a residual SHAPE primitive.

2.4 Why DERIVED-GIVEN-E (not higher, not lower)

• Not FORCED / DERIVED-CLOSED: the count is a rigid integer, but it is read off a selected bundle with $\mathcal{E}$ as input; cross-geometry uniqueness is unproven; the carrier's forcedness over $\mathbb{CP}^2$ uses a 3-aligned tie-break; and $\mathcal{E}$ is not derived. Too much conditioning for an unconditional "derived."
• Not PARTIAL (the SG-4..SG-9 tier): unlike flavor (fitted sector scales) or thresholds (injected unreproduced $\delta$'s), SG-3's central object is a rigid deformation-proof integer that reproduces as a closed-form index, with a clean no-go (APS one-sidedness) for the mirror half. There is no fitted continuous parameter inside the count itself. It sits above the PARTIAL gates.
• DERIVED-GIVEN-E is exactly right: a rigid index (the strong, genuine content) computed with $\mathcal{E}$ as input (the honest qualifier), tied with SG-2 as the two strongest gates in the ledger. This matches the SCOPED_GUT ledger SG-3 line verbatim: "a rigid integer, no dial (Atiyah–Singer) — but computed WITH E as input: it certifies this geometry+bundle yields three families, not that three is unique across all geometries."

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 whether it removes the given-$\mathcal{E}$ conditioning).

3.1 The residual register

3.2 Leverage ranking (attack order)

1. R3 (bundle admissibility / uniqueness) — the single highest-value target: a target-blind bundle-uniqueness theorem is the only route that could remove R1's given-$\mathcal{E}$ conditioning. This is the SHAPE color-rung bundle-admissibility target.
2. R1 (given-$\mathcal{E}$) — conditions every "derived" reading; addressed only via R3 (or refuted as unreachable, which is the honest expectation).
3. R2 (bundle-selected vs carrier-forced; the 3-aligned tie-break) — the live smuggling surface; closing it means forcing 3 without a target-aligned criterion.
4. R4 (spin-$\mathbb{C}$ wording) — cheapest correctness win (refutation already on-disk; correct the structure label to twisted-spin/$\mathbb{Z}_6$).
5. R6 (certificate lint $\to$ machine-checked) — cheapest reproducibility win.
6. R5 / R7 / R8 — disclosed scoping / convention / inheritance items.

The cardinal honest point. R1/R2/R3 are where the "given-E" lives. If a closure path introduces a new target-aligned criterion to force 3 (e.g. "pick the weight that gives three"), it has relocated the input, not removed it — the $\kappa^3/\pi$ falsification test applies in full force here, and the SHAPE suite already flags the +2-dimension forcedness override as the live smuggle. R4/R5/R6/R7 are the non-count residuals (structure label + scoping + reproducibility + convention); closing them improves correctness and machine-reality but does not remove the given-$\mathcal{E}$ conditioning.

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 answer — the $\kappa^3/\pi$ falsification test), the specialist target (theorem to hand off) or the owner artifact (computation/ruling) needed, the math to attempt, and the success ladder (DERIVED-CLOSED rare $\to$ AXIOM-CLOSED likely $\to$ sharper-OPEN $\to$ REFUTED). Dispositions are kept conservative per the closure-campaign norm (default to the more conservative verdict; demote on skeptic-verify).

4.1 R3 — Bundle admissibility / "does $\mathcal{E}$ force the $K_6$ weight?" (the highest-value target)

Why first. Of all SG-3 residuals, only R3 can in principle remove the given-$\mathcal{E}$ conditioning. If the admissible bundle on $K_6$ is unique (forced by the carrier + the $\mathbb{Z}_6$ centre + the hypercharge ledger, with no 3-aligned tie-break and no LEP input), then "3" is forced by the geometry+admissibility alone, and SG-3 moves from DERIVED-GIVEN-E toward FORCED. This is the SHAPE color-rung bundle-admissibility target named in the prompt's closure guidance.

Technique: T10 (selector) + T4 (no-go on alternatives). Run the admissibility selector target-blind over the full space of $K_6$ line/spinor weights compatible with the $\mathbb{Z}_6$ centre and the hypercharge ledger, and ask: is the weight returning $\lvert\chi\rvert=3$ the unique admissible one — or merely one admissible one selected because it gives three?

Named axiom it could reduce to ($\kappa^3/\pi$-clean). The honest target-blind posit references only the carrier + centre + minimality, never the number 3:

AXIOM-MIN-WEIGHT-LIFT (target-blind form): "the admissible bundle on $K_6$ is the minimal-weight twisted-spin ($(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$) lift compatible with the $\mathbb{Z}_6$ centre and the hypercharge ledger; the family count is its Borel–Weil–Bott index."

This is writable with no flavor/generation number in sight — it is a statement about which bundle is admitted, not what its index equals. It PASSES the $\kappa^3/\pi$ falsification test as a statement. The test is then whether it yields $\lvert\chi\rvert=3$ without the 3-aligned tie-break. Honest caution: the corpus already records the nearby A5-actor (SHAPE Lemma 3) attempt — "is $\mathcal{E}_{\rm matter}\oplus\dots$ the minimal actor realization?" — DEMOTED to OPEN because its success criterion is a universal negative ("no admissible competitor supplies a lower-cost realization") that is UNMET (competitor matrix all "Unknown"; NCG un-scored). R3 inherits this hazard: a uniqueness theorem requires ruling out all admissible weights, a universal negative.

Specialist target (to hand off). A bundle-uniqueness theorem: "Given $K_6=SU(3)/T^2$, the $\mathbb{Z}_6$ centre $\ker(\text{centre}\to\mathrm{Aut}(\mathcal{E}))$, and the hypercharge admissibility ledger — and no 3-generation criterion — the minimal-weight admissible twisted-spin lift is unique and its BWB index has magnitude 3." If proven, R3 closes, R2's tie-break dependence falls, and R1's conditioning lifts.

Math to attempt. (1) Enumerate the $K_6$ weights compatible with the $\mathbb{Z}_6$ centre (the Tong-congruence lattice $q\equiv3z_2-2z_3\bmod6$). (2) Apply the BWB index to each; record the family-count map weight $\mapsto\chi$. (3) Check whether minimality (lowest weight) selects a unique weight, and whether that weight's index is $3$ without invoking "three." (4) Sensitivity: does any neighboring admissible weight also give 3 (degenerate) or give 2/4 (excluded only by LEP)? If 2/4 are excluded only by LEP, R3 stays open and R5 is confirmed.

Success ladder. - DERIVED-CLOSED: minimal admissible weight proven unique and its index forced to 3 with no 3-aligned tie-break, no LEP input. Unlikely — the universal-negative obligation (rule out all weights) mirrors the demoted A5-actor Lemma 3. - AXIOM-CLOSED (realistic ceiling): AXIOM-MIN-WEIGHT-LIFT named; the carrier+centre forced, the minimality of the weight posited as a target-blind axiom that is checked to yield 3. Honest endpoint if the check passes. - sharper-OPEN: minimality does not single out a unique weight, or excludes 2/4 only via LEP $\Rightarrow$ "3" stays rigid-given-a-selected-bundle, with the selection conceded. The expected outcome. - REFUTED: a strictly-lower-cost admissible weight gives $\chi\neq-3$ $\Rightarrow$ the minimal-weight axiom is false and the bundle was 3-selected (the good kind of negative — a structure-first datum departing from the target).

Honest disposition: OPEN, with one concrete bounded computation (the weight-enumeration + index map). The realistic ceiling is AXIOM-CLOSED (AXIOM-MIN-WEIGHT-LIFT) or sharper-OPEN; DERIVED-CLOSED is gated on an unreachable universal negative. Do not bank any "minimality" claim that secretly uses "it gives three" as the selector — that is the $\kappa^3/\pi$ failure mode and the SHAPE-suite live smuggle.

4.2 R1 — given-$\mathcal{E}$ ("3" is computed with $\mathcal{E}$ as input)

The obstruction stated precisely. $\chi(K_6,\mathcal{E})=-3$ is the index of a bundle $\mathcal{E}$ that is the SM chiral content. So the index certifies "this geometry+bundle yields three," but it cannot force $\mathcal{E}$ (using $\chi=-3$ to force $\mathcal{E}$ is circular — $\mathcal{E}$ is its input), and it does not establish that three is unique across geometries. This is the binding qualifier of the entire gate.

Technique: T1 (axiom-floor) — name the residual cleanly; closure is via R3 or not at all. R1 is not independently closeable: it dissolves only if R3 (bundle uniqueness) closes (then "3" is forced by geometry+admissibility, not read off $\mathcal{E}$). Absent R3, the honest move is to name the floor:

AXIOM-CONTENT-GIVEN ($\kappa^3/\pi$-clean): "the SM chiral content $\mathcal{E}$ (the matter bundle and its hypercharge ledger) is a SHAPE primitive supplied as input; the geometry computes the family count of $\mathcal{E}$, it does not derive $\mathcal{E}$."

This is writable with no generation number — it is a statement about what is input, not its value. It is essentially already where the corpus stands (REDUCTION_SPINTWIST_C §3(iii): "E remains a residual SHAPE primitive, irreducible-under-known-reductions"; REVIEW_SHAPE_SUITE: the suite bottoms on E, "the un-forced core," and the residue-mislabel-to-fix is precisely that the suite bottoms on "Shape" instead of E).

Specialist target / owner ruling. None new at the math level — R1's only genuine math route is R3. The owner artifact is a labelling decision: ensure every "three families derived" headline carries the given-$\mathcal{E}$ qualifier (the ledger already does; the manuscript §6.4 card says "Claimed certificate pass," and Objection 4 answers "the index is a topological integer ... routed through the selector, not hand-picked," with the honest residue named — "minimality holds only inside the declared search category").

Why this is NOT a relocation. Naming AXIOM-CONTENT-GIVEN does not turn one hard problem into three harder ones; it records the debt in plain sight (the debt = derive $\mathcal{E}$, which is out of scope of scoped-GUT, GAP-class, Gate 11). It is the conservative, honest floor.

Success ladder. - DERIVED-CLOSED: only if R3 closes (3 forced by geometry+admissibility). Unlikely. - AXIOM-CLOSED (where it already sits): AXIOM-CONTENT-GIVEN named; $\mathcal{E}$ conceded as a SHAPE primitive; the gate honestly DERIVED-GIVEN-E. This is the realistic and current endpoint. - sharper-OPEN: if R3's enumeration shows the bundle is degenerate/data-pinned $\Rightarrow$ the conditioning is sharper but unremoved. - REFUTED: not applicable (R1 is a scoping fact, not a falsifiable claim).

Honest disposition: AXIOM-CLOSED at AXIOM-CONTENT-GIVEN; the gate is correctly DERIVED-GIVEN-E and stays there unless R3 closes. given-E $\neq$ derivation of E — this is the load-bearing sentence of the whole dossier.

4.3 R2 — Bundle-selected vs bare-carrier-forced (the SHAPE color-rung finding / the live smuggle)

The obstruction. "3" is read off a chosen weight. On the cheaper carrier $\mathbb{CP}^2=SU(3)/U(2)$ the family count is a continuous bundle-moduli choice — "three by dial." $K_6$ is preferred because its count is a rigid integer, but that preference is purchased by paying +2 dimensions ($\mathbb{CP}^2$ is 4D, $K_6$ is 6D) under a tie-break — anti-fitting / "adjustable counts as fail" — that is target-aligned with the 3-generation count (REVIEW_SHAPE_SUITE defect 4: this overrides the declared lower-dimension Occam rule with a 3-aligned criterion; the suite's live smuggling surface to watch).

Technique: T5 (no tuning to the known answer firewall) + T10 (selector). The firewall question: is the anti-fitting tie-break a genuine prior pass/fail constraint (it was written into the selector before and independently of the 3-generation target), or was it inserted to make $K_6$ beat $\mathbb{CP}^2$ (i.e. to reach 3)?

The firewall, datum by datum:

The decisive point. The carrier choice $K_6$-over-$\mathbb{CP}^2$ is legitimately forced for gauge (Gate 2, abelian-isotropy uniqueness — a data-independent result). The family-count preference, however, leans on the anti-fitting tie-break, and the corpus itself flags this as the live smuggle. The honest split:

AXIOM-COUNT-MUST-BE-INDEX ($\kappa^3/\pi$-clean): "an admissible family count must be a deformation-proof topological integer (an index), not a continuous bundle modulus."

This is writable with no value — it is a type predicate ("the count must be the kind of number an index is"), not "the count is 3." It PASSES the falsification test as a statement. It legitimately kills $\mathbb{CP}^2$'s tunable count regardless of the value — a $\mathbb{CP}^2$ tuned to four fails it just as a $\mathbb{CP}^2$ tuned to three does. This is the genuine, non-target-fitted content of the tie-break, and stating it this way defuses the smuggle: the criterion rejects "three by dial" because it is a dial, not because it is three.

Specialist target. Prove that every clean rigid-integer-count $SU(3)$ carrier of dimension $<6$ either over-produces gauge (like $\mathbb{CP}^2$) or has a moduli-dependent count $\Rightarrow$ $K_6$ is the minimal-dimension carrier with a deformation-proof family index, independently of the value 3. If proven, R2 closes at the type level (the +2-dimension payment is justified by "rigid-integer-count" not "three").

Math to attempt. (1) Compute the family-count map on $\mathbb{CP}^2=SU(3)/U(2)$ explicitly and confirm it is a continuous bundle-modulus (not an index). (2) Verify AXIOM-COUNT-MUST-BE-INDEX rejects it for being a dial, value held abstract. (3) Survey the sub-6D $SU(3)$ carrier list for any rigid-integer-count survivor; confirm $K_6$ is the minimal one. (4) Falsification test: re-run the selector with the 3-generation target masked — does the +2-dimension payment still go through on "rigid-integer-count" alone? If yes, the smuggle is defused; if no, it is confirmed.

Success ladder. - DERIVED-CLOSED: $K_6$ proven the unique minimal-dim rigid-integer-count carrier, value-blind $\Rightarrow$ the +2-dimension payment is justified by type, not target; smuggle defused. Possible (bounded computation). - AXIOM-CLOSED (likely): AXIOM-COUNT-MUST-BE-INDEX named as the target-blind tie-break; the +2-dimension payment conceded as buying rigidity (kind-of-number), not the value 3. - sharper-OPEN: the masked-target selector run shows the payment only goes through because it yields 3 $\Rightarrow$ the smuggle is confirmed and R2 stays open (the honest negative the SHAPE suite warns about). - REFUTED: a sub-6D carrier has a rigid-integer count $\neq3$ that the selector should have preferred $\Rightarrow$ the carrier choice was 3-selected.

Honest disposition: OPEN, with the live smuggle named and a clean target-blind type-axiom (AXIOM-COUNT-MUST-BE-INDEX) that would defuse it — pending the masked-target selector run. The carrier-over-$\mathbb{CP}^2$ choice is legitimately forced for gauge (Gate 2); the family-count tie-break must be re-run target-blind to confirm it is "rigid-integer" not "three."

4.4 R4 — The "spin-$\mathbb{C}$" wording is REFUTED for pure SM (the cleanest correctness win)

The obstruction. The manuscript (§6.4, Appendix E.1, CR4.5, GP.3) calls the family-count object the spin-$\mathbb{C}$ Borel–Weil–Bott index. But pure SM content provably forbids a spin-$\mathbb{C}$ structure: no $U(1)\subset G_{\rm SM}$ gives all-odd Weyl charges (Davighi–Gripaios–Lohitsiri arXiv:1910.11277 Sec.7; reproduced target-blind with a non-vacuous scanner — adding $B{-}L$ makes the all-odd scan succeed). The genuine global object forced by $\mathcal{E}$'s charges is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$ (twist order $n=2$; $(-1)^F$ = the $SU(2)$ $2\pi$ rotation identified with the internal centre; $\mathbb{Z}_6$ via the Tong congruence). Asserting "spin-$\mathbb{C}$" literally would itself be a smuggle (claiming a structure the content forbids).

Technique: T4 (formalize the computed no-go) — pure correctness, no new physics. The refutation is already computed and on-disk (REDUCTION_SPINTWIST_C; SHAPE_M0_DELTASPINC §2). The closure is a wording/structure correction, not a value change.

Named result it reduces to ($\kappa^3/\pi$-clean):

AXIOM-TWISTED-SPIN-GIVEN-E (target-blind, in fact a computed reduction): "the global spin structure on the active branch is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$ forced by $\mathcal{E}$'s charges mod 2; the family count is the index of the associated twisted-Dirac operator on $K_6$, $\lvert\mathrm{Index}\rvert=3$; spin-$\mathbb{C}$ is OBSTRUCTED for pure SM."

This carries no generation number; it references only the SM charges, the centre, and Dai–Freed single-valuedness. It is FORCED-GIVEN-E (REDUCTION_SPINTWIST_C falsifier table: all three conditions PASS — spin-from-E, lands-on-E, E-stays-primitive). Crucially: the value $\lvert\chi\rvert=3$ is unchanged by this correction — a count is not a structure label.

Owner artifact needed. A manuscript wording patch (owner-gated; No status was ever upgraded, nothing applied here): replace "spin-$\mathbb{C}$ index" with "twisted-spin ($(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$) index" in §6.4 / Appendix E.1 / CR4.5 / GP.3, citing arXiv:1910.11277 (DGL) + arXiv:1705.01853 (Tong) + Hsieh–Tachikawa–Yonekura + arXiv:1808.00009 (García-Etxebarria–Montero). The §0.2 on-disk anchors that say "spin-$\mathbb{C}$" are the ones to correct.

Math to attempt. None new — the no-go and the $\mathbb{Z}_6$/twist are computed (scripts cited in REDUCTION_SPINTWIST_C §7). The remaining verification: confirm the twisted-Dirac index on $K_6$ returns the same magnitude 3 as the BWB-on-line-bundle computation did (consistency of the corrected structure with the frozen value). Note: the closure-campaign batch1 flagged the twisted-Dirac spectrum CSV (k6_dirac_spectrum.csv) ABSENT — so a full from-scratch twisted-Dirac re-index is BLOCKED_INPUTS pending that artifact.

Success ladder. - DERIVED-CLOSED: twisted-Dirac index on $K_6$ re-computed = 3, confirming the corrected structure reproduces the value. BLOCKED_INPUTS until k6_dirac_spectrum.csv is mounted. - AXIOM-CLOSED / DISCLOSED-CORRECTED (the realistic, cheap win): wording corrected to twisted-spin/$\mathbb{Z}_6$; spin-$\mathbb{C}$ obstruction disclosed; value $\lvert\chi\rvert=3$ unchanged. This is the cleanest correctness upgrade in the gate and requires only an owner countersign. - sharper-OPEN: if the twisted-Dirac re-index is blocked, the structure is corrected on paper but the value-reproduction-under-the-corrected-structure stays AUDIT. - REFUTED: not applicable (the refutation is of the wording; it strengthens, not weakens, the gate's honesty).

Honest disposition: DISCLOSED-CORRECTED is reachable now (owner wording patch); DERIVED-CLOSED of the value-under-corrected-structure is BLOCKED_INPUTS on the twisted-Dirac spectrum CSV. The family count 3 survives the correction intact — this is a technical-honesty win, not a downgrade.

4.5 R5 — The index-changing deformation excluded by data, not topology

The obstruction. The index $-3$ is deformation-proof inside the declared admissibility class. The neighboring values (2, 4) that an index-changing bundle deformation would produce are excluded by the LEP $N_\nu$ bound ($2.984\pm0.008$), i.e. by data (C2, C4, Appendix E; Objection 4 answer). So "no dial" is topological in-category + empirical at the class boundary — two different kinds of robustness.

Technique: T6 (all-operator-conditional) — scope the claim correctly. No closure path "solves" this; the task is to state the two robustnesses distinctly so neither is overclaimed.

Named principle ($\kappa^3/\pi$-clean):

AXIOM-INDEX-RIGID-IN-CATEGORY "the family count is invariant under every continuous deformation within the declared admissibility class; index-changing deformations across the class boundary are excluded by the LEP $N_\nu$ bound, an external datum."

Carries no value; separates the topological claim from the empirical exclusion.

Action (no new physics). Confirm every "no dial / no fourth family by topology" statement is scoped: deformation- proof in-category (topology) vs at the boundary (LEP). The manuscript already routes the index-changing deformation exclusion to LEP (Objection 4); the residual is keeping that distinction visible wherever "topologically forced" appears.

Success ladder. DISCLOSED-CONSISTENT — already substantially in place (Objection 4 names LEP as the excluder); the residual is a consistency sweep. (Note: closing R3 would make the in-category rigidity also boundary-rigid — so R5 is the dependent of R3.)

Honest disposition: DISCLOSED-CONSISTENT; keep the topological-vs-empirical robustness split explicit. No new physics; keeping the "no dial" claim honestly scoped.

4.6 R6 — The G04_chirality certificate is a consistency lint, not a from-scratch index re-derivation

The obstruction. certificates/G04_chirality/ lints the parity table (survivors = SM set, mirrors projected, counts consistent) against the declared index values $-3$, $(+3,0)$ — under those declared values (CR4.11). It is a consistency check, not an independent re-derivation of the index from the bundle data target-blind.

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

Owner artifact needed. (1) Mount the $K_6$ bundle data + the twisted-Dirac / BWB computation. (2) Re-derive the index from the bundle data target-blind (not read the declared $-3$). (3) Confirm it equals the frozen $-3$ and that the APS boundary count re-derives $(+3,0)$ from the parity ledger ac4d2df3e708. (4) Re-hash R1.3/R1.4 and confirm the meta-hash recomputes to a5b1e6f9d951. Caveat: the from-scratch twisted-Dirac re-index shares R4's blocker (k6_dirac_spectrum.csv ABSENT) — so this is BLOCKED_INPUTS until that CSV is supplied; the BWB-on-line- bundle leg is symbolically re-derivable now.

Math to attempt. None new — execution + verification. Fail-closed: if the index cannot be regenerated from the bundle data, Gate 4 downgrades per the §5.3.7 falsifier (count $\neq3$ / count moduli-dependent / mirror survives).

Success ladder. BLOCKED_INPUTS until the bundle-data + spectrum CSV are mounted $\to$ then VERIFIED (index regenerates target-blind; "Claimed certificate pass" becomes machine-real beyond a lint) or REFUTED (the index fails to regenerate $\to$ downgrade). High value-per-effort because it converts an asserted index into a machine-checked one with no new physics — but gated on the same missing twisted-Dirac artifact as R4.

Honest disposition: BLOCKED_INPUTS (twisted-Dirac spectrum CSV absent) for the from-scratch re-index; the BWB-on-line-bundle symbolic leg + the APS parity-table leg are re-derivable now. Mount the CSV to reach VERIFIED.

4.7 R7 — The chirality/sign convention bit (Pin$^\pm$) is unpinned

The obstruction. The magnitude 3 is robust; the sign $-3$ (chirality orientation — "left-handed, not right-handed") rides the same Pin$^+$/Pin$^-$ / convention bit that leaves BG-10's $e^{\pm i\pi/4}$ unfixed (SHAPE_M0_DELTASPINC §4; HANDOFF_SPECIALIST_3: $\sigma_\nu=+1\to e^{i\pi/4}$ or $\sigma_\nu=-1,-3\to e^{-i\pi/4}$ — not pinned by the frozen record). The family count is convention-independent; the chirality label is convention-stated.

Technique: T1 (axiom-floor), shared with BG-10/SHAPE. Name the bit:

AXIOM-CHIRALITY-ORIENTATION ($\kappa^3/\pi$-clean): "the orientation/Pin bit fixing the sign of the index as left-handed (rather than right-handed) is a single discrete convention choice on the $S_Y^{\,1}/\mathbb{Z}_2$ reflection; the family-count magnitude is independent of it."

No value; a discrete-bit statement. The attack: show the bit is fixed by the same spin-bordism / Dai–Freed single-valuedness datum that fixes the twisted-spin structure (R4), so chirality is structure-forced, not chosen. If the bit is the default of the twisted structure and that default is correct (left-handed), R7 strengthens; if the default is the wrong sign (as BG-10 Rule A's default departs from its target), that is a structure-first datum worth recording.

Specialist target. Hand the spin-bordism specialist: "Is the chirality-orientation Pin bit on $S_Y^{\,1}/\mathbb{Z}_2$ fixed by Dai–Freed single-valuedness of the chiral-fermion measure, or is it a free convention?" — the same handoff as HANDOFF_SPECIALIST_3.

Success ladder. AXIOM-CLOSED (AXIOM-CHIRALITY-ORIENTATION named; bit isolated) is the realistic endpoint; DERIVED-CLOSED if the bit is shown Dai–Freed-forced. sharper-OPEN if it stays a free convention. Honest disposition: AXIOM-CLOSED; the count is unaffected — this is a chirality-label residual shared with BG-10 and SHAPE, three threads landing on one Pin/convention wall (convergence, not closure).

4.8 R8 — Inheritance: SG-8 flavor and the SG-7 threshold vector depend on "3"

The obstruction. SG-8's generation module $\dim\mathcal{G}_{\rm gen}=3$ is inherited from this index (A1.13; SG-8 §1.1 attack handle "SG-8 inherits SG-3's conditionality"). The SG-7 threshold $\delta$-vector is load-bearing on the count $-3$ (Appendix E / G.3.2: "if the family count were $-2$ or $-4$ the column sums would not reproduce $(+4.8424,-3.1112,-1.7313)$"). A revision of "3" cascades to Gates 5/7/9/10 (CR4.7; CR4.11 downgrade rule).

Technique: none internal — dependency record. This is not a closure target; it records the blast radius so a reviewer sees that SG-3 is the dependency graph's busiest node. The honest action: keep the cascade wired (a Gate-4 downgrade forces Gate-5 $\to$ AUDIT and propagates to 7/9/10).

Success ladder. Not applicable (dependency record). Honest disposition: DISCLOSED; the inheritance is correctly wired. The leverage of SG-3 over the whole programme is high precisely because so many gates consume "3" — which is also why the given-$\mathcal{E}$ qualifier must travel with it everywhere (SG-8 already carries it).

4.9 Attack-plan roll-up

REDUCE-vs-RELOCATE verdict on the plan. The plan does not turn one hard problem into three harder ones. Each path either (a) names a single target-blind axiom that pays a debt in plain sight (R1, R2-type-axiom, R5, R7), (b) attempts a bounded, falsifiable computation that could remove conditioning (R3 weight-enumeration; R2 masked-target selector), (c) is a pure correctness/structure fix already computed on-disk (R4), or (d) is a mechanical owner/CSV task (R6) or a dependency record (R8). The two genuine count-moving attempts (R3, R2) are smaller than the original gate (each is one enumeration/selector run) and both carry an explicit $\kappa^3/\pi$ falsification test so a 3-aligned closure cannot be banked — the SHAPE suite's named live smuggle (the +2-dimension forcedness override) is exactly the trap the falsification test guards against. The honest expected outcome of a full campaign: R4 DISCLOSED-CORRECTED (cleanest win), R1 AXIOM-CLOSED (already there), R5 DISCLOSED-CONSISTENT, R7 + R2-type-axiom AXIOM-CLOSED, R3 sharper-OPEN (universal-negative unmet, mirroring the demoted A5-actor Lemma 3), R6 BLOCKED_INPUTS → VERIFIED if the twisted-Dirac CSV is mounted, R8 DISCLOSED. No DERIVED-CLOSED is promised; the gate would move from DERIVED-GIVEN-E-with-spin-$\mathbb{C}$-wording to DERIVED-GIVEN-E-with-corrected-twisted-spin-structure + a named axiom floor + a machine-verified index — a real correctness/honesty gain, not a promotion. The given-$\mathcal{E}$ qualifier is not removable short of R3 closing, and R3's ceiling is AXIOM-CLOSED.

A. Anchoring & Hardening Map

This is SG-3 run through our internal honesty methodology — the same hardening method the live Gaps & Walls Register applies to the whole programme: for each residual ask gap-vs-wall (is the route known, or is the route itself the problem?), hunt the implicit assumption the residual smuggles in, find the measured invariant the residual must ultimately terminate on, then record the honest disposition and the one concrete move that would harden it. The framework and the disposition vocabulary used here (DERIVED / DERIVED-GIVEN-E / AXIOM-CLOSED / DISSOLVED / SCHEME-ANCHORED / OPEN / BLOCKED / measured-but-irreducible) are defined in, and consistent with, the live register — https://physics.magflowmeters.com/articles/GAPS_AND_WALLS_REGISTER.html. The existing anchor set this gate can terminate residuals on is {ℏ, M_Pl, spectrum-E, α_i(M_Z), y_t, |V_us|}; a residual hardens only when it reduces to one of those (or to a single named, value-free, target-blind posit), never when it reduces to "it gives three."

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

A.2 Gate-level rollup

• Overall disposition. SG-3 is DERIVED-GIVEN-E and stays there. The strong, genuine content — a rigid, deformation-proof integer $\lvert\chi(K_6,\mathcal{E})\rvert=3$ (BWB, closed form) plus a clean no-go for the mirror half (APS $(n_L,n_R)=(+3,0)$) — is real and reproduces; it ties SG-2 as the strongest gate. The honest floor is that every count residual terminates on spectrum-E (the index reads $\mathcal{E}$; the 2/4 deformation is excluded by the measured LEP $N_\nu$), so "three" is a rigid facet of the observed spectrum, not a from-nothing derivation. Of the eight residuals: R1/R2/R3 are WALLs (route-is-the-problem / target-aligned escape / universal negative); R4/R5/R6/R7/R8 are GAPs (one BLOCKED on a missing file). No residual is DISSOLVED, and none promotes.
• Anchors it depends on. Primarily spectrum-E (the SM chiral content $\mathcal{E}$ and its hypercharge ledger), reinforced by the external measured invariant LEP/SLD $N_\nu$ = 2.984 ± 0.008 at the admissibility-class boundary. It does not depend on ℏ, M_Pl, α_i(M_Z), y_t, or |V_us| for the count itself (those enter downstream gates that consume "3"). R7 would need a new named invariant (the Dai–Freed orientation datum) that no existing anchor supplies.
• Single highest-leverage hardening move. R3 — the target-blind bundle-uniqueness computation (enumerate the $\mathbb{Z}_6$-compatible $K_6$ weights, map weight $\mapsto\chi$, test whether minimality alone forces $\lvert\chi\rvert=3$ with no 3-aligned tie-break and no LEP input). It is the only path that could lift R1's given-$\mathcal{E}$ conditioning; its realistic ceiling is AXIOM-CLOSED (AXIOM-MIN-WEIGHT-LIFT), and any "minimality" claim that secretly uses "it gives three" as the selector is the $\kappa^3/\pi$ failure mode and must be killed, not banked. The cheapest correctness win remains R4 (the spin-$\mathbb{C}$ → twisted-spin wording patch, value 3 intact). Ceiling: serious candidate, NOT validated. No status was ever upgraded; frozen branch dcc66f1b2685 / a5b1e6f9d951 READ-ONLY; nothing applied, nothing deployed.

🎯 Target anchor(s) for this gate

In the terminate-on sense (every residual must reduce to a measured invariant or a single named, target-blind posit — never to "it gives three"), this gate's target anchor is the observed spectrum E itself: the family count "3" — and the rigid integer $\chi(K_6,\mathcal{E})=-3$ that carries it — is a facet of E ($\chi$ is read with $\mathcal{E}$ as input), reinforced at the admissibility-class boundary by the external measured invariant LEP/SLD $N_\nu = 2.984 \pm 0.008$ (which excludes the index-changing deformation to 2/4). Honest status: spectrum-E is measured-but-irreducible (the index reads it; it cannot derive it — given-E $\neq$ derivation of E); the would-be deeper target "3 across all geometries / bare-carrier–forced" has no independent witness and stays OPEN (R1/R2/R3); the in-category "no-dial" rigidity is SCHEME-ANCHORED to the declared search category (AXIOM-INDEX-RIGID-IN-CATEGORY), and the orientation/chirality-label bit awaits a new named invariant (Dai–Freed, AXIOM-CHIRALITY-ORIENTATION). The count's magnitude is genuine and reproduces; its forcedness across geometries is not anchored. No status was ever upgraded.

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
• §6.4 Gate-4 card (chirality / no mirrors / family count; binding status); §6.12 falsification map (Gate-4 row); §6.13 certificate summary.
• §5.3 narrative module (Gate 4); §3.4 topological-count machinery; §4.6 selector dimension-payment ($\mathbb{CP}^2\to K_6$, "+2 dimensions for forcedness"); §4.9 anti-fitting firewall ("three by dial" fails).
• The Strongest Objections block — Objection 4 ("the three-family count is asserted by choosing the geometry"; answer: topological integer routed through the selector, residue named — minimality holds only inside the declared search category; LEP $N_\nu$ excludes the index-changing deformation).
• Appendix E (Chirality Closure — authoritative): E.1 mechanism, E.2 three generations, E.3 mirror ledger, E.6 certificate; Appendix CR module CR4 (reader companion, CR4.1–CR4.13).
• Appendix GP.3 (the counting theorems — Atiyah–Singer, APS, Borel–Weil–Bott); B1/B2 (selector, category-relativity B2.0.3.4); C2 ($K_6$ dossier), C4 ($S_Y^{\,1}/\mathbb{Z}_2$ dossier), A2.2 (bundle ledger), A1.8 (parity table), A1.13 (generation module); Appendix R0 freeze records.
• Paper IV, TOE.html — https://physics.magflowmeters.com/articles/TOE.html (downstream cross-references only; does not touch SG-3).

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

5.2 Corpus locations (authoritative inputs to this dossier)

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

Branch dcc66f1b2685 · manifest meta a5b1e6f9d951 · spin-$\mathbb{C}$/twisted-spin bundle data on $K_6$ (R1.4) 0fd19c9ae0c1 · $\mathbb{Z}_2$ orbifold + boundary parity ledger (R1.3) ac4d2df3e708. Carrier $K_6=SU(3)/T^2$, $\chi(K_6)=6$, and the index value $\chi(K_6,\mathcal{E})=-3$ / APS $(n_L,n_R)=(+3,0)$ are frozen R1.3/R1.4 outputs. No new hash is introduced by this dossier; nothing is mutated. Missing artifact (blocks R4/R6 from-scratch re-index): k6_dirac_spectrum.csv — ABSENT (closure-campaign batch1; twisted-Dirac spectrum not on disk).

5.4 External literature (cited at the corpus's own citation level, for R4/R7)

• Davighi, Gripaios, Lohitsiri, "Global anomalies in the Standard Model(s) and Beyond," JHEP 07 (2020) 232, arXiv:1910.11277, Sec. 7 — pure SM has no spin-$\mathbb{C}$ without an extra gauged $U(1)$ (the R4 refutation).
• Tong, "Line Operators in the Standard Model," JHEP 07 (2017) 104, arXiv:1705.01853 — $\mathbb{Z}_6$ = maximal trivially-acting centre; congruence $q\equiv3z_2-2z_3\bmod6$ (the R2/R4 centre kernel).
• Hsieh, Tachikawa, Yonekura, "Anomaly of the Standard Model" — twisted $\mathrm{Spin}\times G_{\rm SM}/\mathbb{Z}_q$; $(-1)^F$ in the gauge centre.
• García-Etxebarria & Montero, "Dai–Freed anomalies in particle physics," JHEP 08 (2019) 003, arXiv:1808.00009 — Dai–Freed / spin-bordism fermion-measure (the R4/R7 single-valuedness datum).
• Allanach et al., arXiv:2111.04148 + anomaly-free atlas JHEP 02 (2019) 082 — infinitely many anomaly-free chiral $U(1)$ extensions ⇒ anomaly-freedom fixes neither content nor generation number (the R1/T3-REFUTED basis).

Closing honest statement

SG-3 is DERIVED-GIVEN-E — and, among the ten gates, it ties SG-2 as the strongest, because its central object is a rigid, deformation-proof integer ($\lvert\chi(K_6,\mathcal{E})\rvert=3$, computed in closed form by Borel–Weil–Bott) plus a clean no-go for the mirror half (APS one-sidedness $(n_L,n_R)=(+3,0)$ on the $S_Y^{\,1}/\mathbb{Z}_2$ fold). The magnitude 3 is the family count; the sign and the one-sidedness are chirality; there is no dial. That is genuine, defensible, and reproduces (the rep-theory engine — $m_0(p,q)$, the $\mathbb{Z}_6$ Tong congruence — regenerates target-blind). Its honest open surface is equally clear: the index is computed WITH $\mathcal{E}$ as input (given-E — it certifies this geometry+bundle yields three, not that three is unique across geometries, and it cannot force $\mathcal{E}$, which stays a SHAPE primitive — T3 REFUTED); the count is bundle-selected, not bare-carrier-forced ($\mathbb{CP}^2$ gives "three by dial"; $K_6$'s forcedness is bought by paying +2 dimensions under a 3-aligned tie-break — the SHAPE suite's named live smuggle); the admissible bundle is data-/centre-fed (LEP $N_\nu$ excludes the index-changing deformation), not internally proven unique; and the literal "spin-$\mathbb{C}$" wording is REFUTED for pure SM (Davighi–Gripaios–Lohitsiri — the genuine forced object is the twisted $(\mathrm{Spin}\times G_{\rm SM})/\mathbb{Z}_6$, FORCED-GIVEN-E, with the value 3 intact). The attack plan reduces these to named, target-blind axioms (AXIOM-MIN-WEIGHT-LIFT, AXIOM-CONTENT-GIVEN, AXIOM-COUNT-MUST-BE-INDEX, AXIOM-TWISTED-SPIN-GIVEN-E, AXIOM-INDEX-RIGID-IN-CATEGORY, AXIOM-CHIRALITY-ORIENTATION) and bounded falsifiable computations (the weight-enumeration index map; the masked-target selector run) — with the $\kappa^3/\pi$ falsification test guarding R2/R3 so the 3-aligned forcedness override cannot be dressed as a derivation. The realistic ceiling is a corrected (twisted-spin) structure + a named axiom floor + a machine-verified index — the given-$\mathcal{E}$ qualifier is not removable short of a bundle-uniqueness theorem (R3), whose own ceiling is AXIOM-CLOSED. No status was ever upgraded; frozen branch dcc66f1b2685 / a5b1e6f9d951 READ-ONLY; given-E $\neq$ derivation of E; selection $\neq$ derivation; nothing applied, nothing deployed.

Dossier built 2026-06-24. Our geometry (13D $K_6$ branch) only. Common material referenced to the published website source-of-truth, not duplicated.