# (5) WHAT TO ATTACK FIRST — THE CORPUS'S OWN WEAKEST-LINKS-FIRST POSTURE

Every manuscript prints its top objections / weakest links / open rows **up front**, routed to
the top for a hostile reviewer, each with a falsifier or named gate attached. This is the
series' central design choice: *find the weak link, and finding the first failing link is
sufficient grounds to downgrade the manuscript-level claim.* **DRAFT — pending countersign (see
the referee-aid readme §0).**

A referee who wants to break the series efficiently should start here. The binding rule across
all docs: **a required gate may never be closed by narrowing scope, and an excluded sector may
never be leaned on to close anything.**

---

## A. TOE_FINAL (capstone) — "Known weakest links (routed first, for a hostile read)"

1. **Gap-04 closure is decision-grade, not certificate-grade.** §II states exactly what the
   countersign cleared, by what margin (35× weakest variant / 2.4×10³ central), and what stays
   open inside the gap (c_loop; a non-blocking magnitude lane).
2. **The Λ verdict is a computed honest FAIL.** The theory's own deepest ⊕-layer
   chamber-cancellation claim was tested and **refuted** (§III: I2 supertrace honest-FAIL + I3
   theorem-refutation). **Λ remains Weinberg-open;** no observed-Λ comparison is performed
   anywhere.
3. **The inflation numbers ride two PENDING countersigns and one convention gate (K13-2).** §IV
   carries the full four-branch table rather than one flattering band.
4. **BG-10 is OPEN at 0/4 fixes.** No baryogenesis number is claimed. A first-pass closure of
   2026-06-04 was withdrawn the same day after a ten-agent audit.
5. **Strong-CP θ̄-smallness is honestly underived** (§V stanzas); nEDM probes it continuously.

**The capstone truth table (§V) distribution — read the shape, not any one row:** of 18 gaps —
**2 closed-by-computation** (Gaps 07 arrow-of-time PASS, 17 reproducibility CERTIFICATE), **1
closed at decision grade** (Gap 04), **5 conditional behind a named gate** (Gaps 08, 09, 11, 16,
18), **10 precisely-OPEN** with a falsifier or recorded non-claim attached (Gaps 01, 02, 03, 05,
06, 10, 12, 13, 14, 15). Counting sub-items: 48 total — 22 closed-by-computation, 3 at decision
grade, 9 conditional, 14 precisely-OPEN. A table placing all 18 rows in the first state would be
the thing to distrust.

**The two deliberate Clay walls (kept standing, not failed gates):** Wall 1 — Yang-Mills mass gap
(Gap 02, CANDIDATE on C1–C4 structural compatibility, **not** a nonperturbative proof). Wall 2 —
UV quantum gravity at field grade (Gap 01, single named missing object = the a₆ Seeley-DeWitt
trace at d=13).

---

## B. Paper 1 — GUT — "Known Weakest Links Before Review" (10 named, attack-first)

The manuscript lists its ten most vulnerable points so reviewers attack them first; if any one
fails, the manuscript downgrades from claimed certificate closure to partial mechanism:

| Weak link | Why vulnerable | Where to inspect |
|---|---|---|
| F⁺ flavor chamber | Could read as compressed Yukawa fitting unless calibration inputs/outputs/freeze timing are airtight | §7, App. I/J/K, R0 |
| Threshold unification | Easy to overfit unless spectrum, regulator, comparison scale, freeze records are reproducible | App. G, R0 |
| Higgs protection | Wilson-line protection must prevent the high-scale sensitivity, not just rename the Higgs | App. H, C9 |
| Proton safety | Projector identity must eliminate the *physically relevant* dangerous operator class | App. L, C10 |
| Stabilization | Must distinguish actual moduli stabilization from admissibility restriction | App. F |
| Search category | Minimality is only meaningful inside the declared search category | App. B1, B2 |
| Three-layer architecture | ⊕ and ⊗ must be structural, not bookkeeping | §2B, App. B2, A2, A3 |
| Term-level necessity | Each retained term needs a local failure-if-removed certificate | App. C |
| Freeze discipline | Every comparison-relevant object frozen before comparison | R1, R0 |
| Claim language | "Closure" = certificate closure under stated assumptions, not final proof of nature | Abstract, §1, §10 |

**The GUT's "limited time — attack the weakest links first" priority route (one hour):**
1. **App. L (Proton safety)** L.2a/L.2b/L.10 — does Π_q M Π_ℓ = 0 eliminate the *physically
   dangerous* operator class, not only a narrowed declared class?
2. **App. I (Quark flavor)** I.0a/I.10/I.11 — does the F⁺ chamber disguise Yukawa fitting?
   (anchor-vs-output ledger; anti-fitting audit)
3. **App. G (Threshold unification)** G.0/G.10 — do thresholds smuggle hidden free parameters?
4. **App. H (Higgs protection)** H.0/H.10 — does Wilson-line protection eliminate the *relevant*
   correction class, or only rename the Higgs?
5. **App. R0 (Freeze)** R0.10/R0.11 — is the freeze record genuinely executable and
   content-addressable?

**Excluded sectors (stated as a strength, fenced by Gate 11):** quantum gravity, full cosmology,
dark matter, dark energy, baryogenesis, strong CP — **excluded, NOT failed gates;** a required gate
cannot be closed by relocating its difficulty into the excluded row.

---

## C. Paper 2 — Forces — "Known weakest links (stated before review)" (§0.2)

- **WL-1 — Einstein-frame conditionality (gravity).** The headline M_Pl² = M₁₃¹¹·Vol(X) is honest
  only after the volume modulus is frozen (authority external: Paper I Gate 6 / App. F). If the
  modulus stayed dynamical the gravity interface self-downgrades to scalar-tensor. A *declared
  dependency on Paper I*, not closed here.
- **WL-2 — Quantum self-consistency of the compactification background is inherited, not proved
  here.** Governed by the corpus (TOE_FINAL Gap-04 discharge + Λ verdict), not this paper.
- **WL-3 — All gate certificates are external.** Every "frozen/anomaly-free/certificate" statement
  resolves in Paper I; this paper re-closes no gate. If a Paper I gate downgrades, consuming rows
  here downgrade automatically.
- **WL-4 — The strong sector's confinement/χSB/strong-CP edge is open.** SU(3)_c recovery is an
  interface claim; confinement, the mass gap, and strong CP are scoped out and *not* used as
  support. Strong CP remains honestly underived in the corpus.

**Attack-first target (§0.1):** the composite map π_QED ∘ π_weak (C-FF8). If that factorization is
not a real map from the frozen 13D active geometry down through the electroweak EFT, the routing
collapses. **Fastest falsification entry (no tools):** the by-hand anomaly witness
3·(1/6) − 1/2 = 0 (§8.6.4) — if the hypercharge lattice were off by one step it fails and the
charge table breaks with it.

---

## D. Paper 3 — Quantum — the AUDIT rows printed on the cover (Σ = AUDIT)

**The headline IS the weakest link.** Σ = min over 17 gates = **AUDIT**, by the min-rule (no strong
result can hide a weak one). The **8 AUDIT rows** that hold the aggregate down — *routed to the top
so a hostile reviewer finds them first* — are:

| Gate | Name | What raises it (the open obligation) |
|---|---|---|
| UQF-4 | BRST/BV nilpotency and anomaly closure | Phase-4 anomaly lift under 4D descent |
| UQF-7 | Fermion quantization | Phase-3 no-light-mirror verification |
| UQF-9 | UV-completion certificate | non-Gaussian fixed point at declared truncation (global open problem; committed to the asymptotic-safety route) |
| UQF-10 | Quantum compactification stability | blocked by UQF-9 (loop control of moduli potential) |
| UQF-11 | Nonperturbative QCD | named lattice / partner-community program (**Yang-Mills mass gap = Clay Millennium problem**) |
| UQF-12 | Strong CP | nEDM falsifier on the boundary-parity candidate (LOW confidence) |
| UQF-14 | Unitarity / causality / locality audit | graviton-sector unitarity above cutoff (UQF-9) |
| UQF-15 | No-hidden-sector force leakage | §4 ledger exhaustiveness + Phase-4 audit |

Raising Σ to CANDIDATE is the declared Phase-7 target and requires exactly those 8 rows; raising it
to CERTIFICATE requires the named partner-community executions. **No sentence in the paper outranks
this table** — the min-rule makes self-flattery a structural drafting bug.

---

## E. THE STRONGEST OBJECTIONS, ANSWERED (each doc carries a steelman section)

Each manuscript states the five strongest hostile objections **at full force first**, then answers
each by pointing at the in-document claim → mechanism → certificate → falsifier chain. The
recurring five (TOE_FINAL / GUT versions):

1. **"More outputs than inputs is a magic trick."** → answered by the registered blind-negative
   control PCM-12 + the no tuning to the known answer firewall; outputs are a topological index and frozen
   chamber values with no spare dial.
2. **"Decision-grade is just 'not proven.'"** → conceded and routed to the top (weakest link #1);
   the operative falsifier was passed 35× / 2.4×10³, and the residue (c_loop) is named and proved
   unable to threaten the well.
3. **"You failed on Λ / you excluded the hard sectors."** → the theory does not claim Λ; it
   refuted its own cancellation mechanism and published it; excluded sectors are stated as a
   strength, fenced by a binding rule.
4. **"Almost everything is OPEN/CONDITIONAL — this is a research programme, not a theory."** →
   conceded from the doc's own counts; every OPEN item carries a falsifier and an owner, so "open"
   = precisely specified and routed, not vague.
5. **"Self-correction is theater."** → the self-catches are directional and public against an
   append-only record (each made a claim *weaker*); the mechanism that caught them is still running,
   in two independent domains (physics + QC engineering).

(Paper 2 and Paper 3 carry domain-specific versions — e.g. Paper 3's lead objection is the
circularity of a self-graded gate stack, answered by "tiers are earned, not declared; 8 rows sit
at AUDIT.")

---

## F. THE NAMED FALSIFIERS ON A CALENDAR (TOE_FINAL §IV world timeline)

Between now and ~2035 the world hands the series multiple independent chances to die — each a
published bound with its date:

| When | What arrives | What it falsifies |
|---|---|---|
| 2027 | m_W | the corpus m_W comparison row |
| 2030+ | LiteBIRD r | the σ-inflaton: r outside the four-branch union [3.5, 36]×10⁻³ falsifies; the branch separation makes the convention an observable |
| ~2035 | τ_p reach | the pre-registered dim-6 QQQL proton-decay falsifier |
| Ongoing | nEDM | θ̄ (Gap 03): θ̄ > 10⁻¹⁰ makes the underived status a refutation |
| Any time | DM direct detection | the minimal Gap-11 candidate (null-forecast σ_SI ~ 10⁻⁵⁵–10⁻⁶⁰ cm²; any signal is fatal) |
| On completion of 4 fixes | η_B | BG-10: η_B outside [6.0, 6.2]×10⁻¹⁰ at 3σ falsifies |

**The referee's takeaway:** the series does not ask to be believed. It routes its own weakest links
to the top, attaches a falsifier or named gate to every open item, and names the experiment and the
year that could end each claim. *Finding the first failing link is a complete, sufficient review.*

---

<!-- contact-footer-cb -->
**Chris Bergstrom** &middot; cbergstr@gmail.com &middot; physics.magflowmeters.com