What this is. A plain-language companion to the technical "Minimal-Shape Suite." It carries the same honest conclusions as the technical version, in words a careful non-physicist can follow start to finish. It proves nothing new, changes nothing, and certifies no new number. It argues a status — and it argues it carefully. The frozen object it discusses (the "13-dimensional geometry," labeled
dcc66f1b2685) is read-only. Nothing here touches it. Nothing in this document gets upgraded from "we think" to "we proved." For the program-wide scoreboard — all 33 requirement-gates at closed endpoints, none open, 0 of 33 physics-closed — see the live gate ledger and the Shape deep-root record; the question examined here — absolute minimality — is the one claim the program deliberately does not make.
There is a tempting sentence that a less careful version of this work would put right at the top:
"We proved the shape of the universe is the only one possible — it's forced."
We will not write that sentence, anywhere, in any disguise, because it is not true. The honest claim is weaker, and the whole point of this document is the gap between the two. Here is the honest claim, in one breath:
Among the candidate shapes we actually wrote down and checked, under a simplicity rule we fixed in advance, the frozen geometry is the simplest one that does the whole job. No serious rival currently beats it. And any workable theory of our world has to carry at least three basic "parts." But none of this forces the shape — it selects it from a menu. Choosing the best item on a menu is not the same as proving it's the only dish that could ever exist.
That distinction — selecting versus forcing — is the entire story. Everything below is the careful defense of what we genuinely earned, and the equally careful confession of what we did not. At the very end there is a real fork in the road, left genuinely open: either the shape stands, or someday something simpler shows up and the shape quietly folds into it. We do not pretend to know which.
This document is written to survive a hostile reading — the kind of reviewer who trusts nobody, who hunts for the place where "we selected it" got dressed up as "we proved it," and who rejects any grand "nothing else could ever work" claim on sight. Where an earlier review already caught real weaknesses in this work, we repeat those weaknesses here in plain sight rather than hiding them.
The word "minimal" is doing dangerous double duty. People use it for two completely different claims, and pretending they're the same is exactly the trap.
Claim A — "Forced" (the strong one). Some deeper principle makes this shape unavoidable. Anyone who accepts the principle has to accept the shape. To knock this down, you'd only need to find one example where the principle holds but the shape doesn't.
Claim B — "Cheapest on the menu" (the honest one). Out of the candidate shapes we listed, and judged by a simplicity rule we set in advance, this one is the cheapest that still gets the full job done. To knock this down, you'd find a cheaper candidate, on the same menu, that also does the full job.
The difference is the word "all." Claim A quantifies over every conceivable structure in the universe. Claim B quantifies over the menu we drew. We earn Claim B (within the menu). We do not claim, and have not earned, Claim A.
Why does the menu version still count for something? Because it isn't vibes — it's a precise, checkable, falsifiable statement: given this list of candidates, this list of requirements, and this simplicity ranking, the frozen shape is the winner. The honesty is in never quietly dropping the "given this list." We always say which menu we're standing on, and we tell hostile reviewers exactly where to attack it: don't attack the winner, attack the menu and the ranking rule.
The hard rule we live by: a search that returns one winner feels exactly like a proof. You run the machine, one option survives, and it's tempting to write "therefore it's forced." But "only one survivor" is only-one-survivor relative to the menu and rule you chose. Change the menu, and the survivor can change. A genuine proof has no such dependence — it survives no matter how big you make the universe of options. So: a unique survivor is always reported as "the selected winner," never as "forced." The moment a sentence says "therefore the shape must be...," it is wrong unless the "must" is openly tied to "...within this declared menu."
The shape is not picked by scanning every possible geometry and admiring one. It's picked by a funnel that runs the other way around: start from the known facts about the particles and forces we actually observe, then ask which candidate shapes survive every required test. The tests are things like: does it reproduce the known forces? the right electric charges? left/right-handedness done correctly? three families of matter? does it stay internally consistent (no fatal contradictions)? does it protect the proton from decaying too fast? and so on — ten tests in all.
The crucial ordering: simplicity is applied last, and only to candidates that already passed every test. This is what stops "minimal" from collapsing into "the cheapest thing that barely does anything." The cheapest object imaginable is empty space with nothing in it — and it's thrown out immediately, because it fails the very first real test (it produces no forces). Cheap but incomplete is disqualified, by rule.
The menu is declared before the comparison, and then frozen with a fingerprint (a hash). This matters enormously for an honest result:
And there's an explicitly banned move, named out loud so a reviewer can check it didn't happen: you may not shrink the menu mid-contest to delete a rival that was beating you. ("Compare against everything... then quietly erase the competitor that won.") That's the classic cheat, and it is forbidden by a written rule.
Here's the heart of it. The simplicity rule (Occam's razor, in the trade) has one absolute restriction stated up front:
You may remove a part of the structure only if the building still stands once it's gone.
In other words, the machine does not select the simplest theory. It selects the simplest theory that is still complete — and "complete" is decided by the ten tests, not by taste. After that one ironclad rule, the tie-breakers prefer: fewer dimensions, fewer free inputs you have to choose by hand, fewer independent knobs, one mechanism per job (no doing the same job twice), and no decorative parts that serve no test.
Boiled down:
Completeness beats simplicity. Simplicity only gets to rank the options that are already complete.
What makes this a real rule and not a convenient excuse is that it cuts in both directions — and we show both:
A rule that both keeps an expensive part and pays for extra dimensions — driven by a single written priority — is behaving like a genuine constraint, not like something reverse-engineered to reach a desired answer. If it only ever cut toward the answer we wanted, that would be a red flag. It cuts both ways.
The obvious suspicion: "'Completeness beats simplicity' sounds like a license to keep whatever you like, just by declaring it 'required.'" The defense is that "required" is not a matter of taste here — it's a yes/no outcome of a re-runnable test. A part is "required" if and only if removing it makes a specific named test fail, and you can watch it fail by re-running the machine. Delete the flavor chamber, re-run, watch the flavor test break. That's a checkable fact, not an opinion. A taste-based "required" would be a cheat. A test-based "required" is a razor. This one is test-based.
There is exactly one place in this whole story where the rule makes us pay more (two extra dimensions) instead of less. A hostile reviewer will go straight there and say: "They reverse-engineered a special bonus for 'forcedness' so that the answer would come out with three families of matter." So let's work it completely, in the open.
Two internal shapes both reproduce the strong-force symmetry exactly. Both pass the early tests — and both, it turns out, deliver a rigid whole-number family count. They differ at exactly one structural test — does the shape hand back exactly the observed forces, and nothing extra? — and only there:
| Candidate | Dimensions | Reproduces the symmetry? | How many families does it give? | Simpler? | Hands back exactly the observed forces? |
|---|---|---|---|---|---|
| The cheaper shape (call it CP²) | 4 | Yes, exactly | A rigid whole number too — but its mounting bracket is not inert | Yes (2 fewer dimensions) | No — its non-commutative bracket turns gauge-active: it over-produces forces we don't observe, or traps the weak/hypercharge pieces inside color (built end-to-end, it broke exactly there) |
| The frozen shape (K₆) | 6 | Yes, exactly | Three — locked in, no dial | No (costs 2 dimensions) | Yes — its bracket is inert (commutative): exactly the observed forces come out, nothing extra |
Here is the whole crux in one sentence: the cheaper shape's mounting bracket is not inert — it switches on extra forces nobody has ever observed. Every candidate in this family hangs on the strong force's symmetry by a "mounting bracket" — the part of the symmetry the shape holds fixed. The frozen 6-dimensional shape's bracket is commutative (inert), so exactly the observed forces come out and nothing else. The cheaper 4-dimensional shape's bracket is non-commutative, and a standard, long-known rule of this construction promotes a non-commutative bracket into additional, unwanted forces in the resulting world — or, said another way, it traps the weak and hypercharge pieces inside the color structure. And this is not an argument on paper: the program built the cheaper shape end-to-end, as an honest stress-test, and watched it break at exactly the predicted spot.
An honesty note belongs right here, stated as history: an earlier version of this comparison leaned on a different — and unsound — argument, that the cheaper shape's family count was a tunable dial you could set to anything. The program itself caught that argument and retired it: the cheaper shape's family count is also a rigid whole number (a topological index), so counting families alone cannot separate these two candidates at all. The replacement argument above is stronger, not weaker — it is structural, all-or-nothing, and completely blind to the target: the cheaper shape fails whether or not three families are ever mentioned. The program-wide anti-fitting discipline (every adjustable-after-the-fact quantity is barred; inputs declared and frozen in advance) still stands everywhere — but the deciding move here is the cleanness of what comes out, not a dial.
So, step by step, with nothing hidden:
This is exactly the same operation as keeping the "expensive" flavor chamber: the cheaper option is incomplete, so the dearer complete option wins. It is not a special prize handed to the 6-D shape for being "forced."
You'll sometimes see this glossed casually as "the selector pays two extra dimensions for forcedness." That phrase is exactly the one a hostile reviewer pounces on, and rightly so. The honest reading — the one the manuscript itself now states — is this: the cheaper 4-D shape was dropped because, built end-to-end, it breaks a required test outright — it cannot hand back the observed forces cleanly. That makes it incomplete. The dearer 6-D shape is kept because it's the cheapest complete option. "Forcedness" is not a new bonus being chased; it's just another name for "passes the test without breaking."
Objection (hostile reviewer): "You built the rival yourselves and you graded it yourselves — and the requirement list it broke against was written by people who already knew which forces, and how many families, the world has. However mechanical the elimination step, the test list itself is aimed at the answer. That's loading the dice one level up."
This is the strongest form, and it deserves a straight answer in three parts.
Part 1 — Is the elimination step itself rigged? No. The move that removes the cheaper shape is structural and target-blind: the shape fails on the cleanness of the forces it hands back — an all-or-nothing feature of its own construction, predicted in advance and then confirmed by the end-to-end build. The failure happens whether or not three families are ever mentioned. So this isn't a bespoke rule conjured to kill one rival. On the narrow question of a rigged elimination, this part of the objection is answered.
Part 2 — Does the rule only ever cut toward the answer? No. The same rule keeps an expensive part (the flavor chamber) and pays for expensive dimensions (the 6-D shape), and elsewhere it cuts cheaper shapes when they're redundant. A dice-loaded rule would only ever push toward the target; this one sometimes costs you simplicity to preserve completeness. That's a constraint behaving like a constraint. On the narrow question of smuggling, this part is answered too.
Part 3 — The part that is NOT answered, and we say so. The objection has a true residue that no amount of "the elimination is target-blind" can erase: the requirement list bottoms out on the fact that the world has three families. "Three families, locked in" is only a requirement because we are matching the three families we actually observe. If the world had two families, the test would demand a locked-in two. So the three-families count is an input — a fact about which particles exist — not something the geometry derives. Which means: the elimination of the cheaper shape is a clean, fair — indeed target-blind — selection step, but the contest it lost was staged against the world we actually observe. It is not, and we do not claim it is, an explanation of why three. That residue stands, and we hand it to the reviewer rather than papering over it.
So the bottom line on this contested case: it is earned as a selection move, and it is explicitly not a derivation of how many families exist.
To show this isn't a tale told once to reach a convenient answer, note the funnel runs the ordinary (cut-toward-cheaper) direction elsewhere too, by the same rule: a 3-sphere shape is dropped in favor of a 2-sphere ("one extra dimension, no test served"), and a more elaborate twisted-torus is dropped in favor of a simpler circle-fold ("everything it offers is already supplied more cheaply"). Those are ordinary simplicity cuts. The 6-D-over-4-D case is the unusual direction (paying more), run by the same machine with the completeness rule switched on. A reviewer who accepts the ordinary cuts as fair Occam reasoning has to accept that the unusual one is the same machine — with the only remaining thing to contest being the inputs the requirement list is matched against (which particles and forces exist), which we already concede are unforced.
Everything in sections 2–3 lives inside the menu. The next result reaches for something broader — a statement about any workable theory of our world, not just shapes of this particular family. The claim:
Any theory that meets the real physical burden must contain functional versions of three basic parts: - a Stage — somewhere for things to live and propagate (the geometry / space part); - a Rulebook — what's allowed and what isn't (the selection / consistency part); - Actors — the actual stuff: matter, forces, the Higgs, the proton (the content part).
A rival theory doesn't have to use our words or our notation. But it must carry these three jobs somehow — and if it tried to drop one, it would fail some part of the physical burden.
The earned part: you need at least three parts. The argument is a simple three-way contradiction — drop the Stage and there's nowhere for anything to exist; drop the Rulebook and nothing constrains what's allowed; drop the Actors and there's no matter or forces. The frozen shape has exactly three parts, sitting right on the floor.
But here's the careful limit: this says "at least three," not "these particular three are the cheapest." It rules out any rival that tries to win by having fewer than three parts. It says nothing about whether our filling of those three parts is the simplest possible filling. (That harder question is the open one — see section 6.) There's also a guard against a cheap dodge: you can't win by fusing parts to "have fewer pieces," because fewer pieces isn't fewer jobs — if you fuse two jobs into one piece, that piece is just secretly doing two jobs.
An earlier hostile review caught two real problems here, and we state them plainly.
Weakness 1 — the argument is close to circular. The three "parts" are defined as restatements of the very requirements they're supposed to follow from. So "the requirements imply you need a Stage" is uncomfortably close to "these requirements imply a name for these requirements." That's true, but it's nearly saying the same thing twice. What survives: the count — that you genuinely can't do the job with fewer than three distinct jobs — carries a little real content. What does not survive: you cannot use this floor as a springboard to any grand "nothing simpler could ever exist" claim. A near-tautology can't be a gateway to a sweeping universal.
Weakness 2 — the "physical burden" smuggled in some house rules. The list of requirements was billed as universal — true of any workable theory. But two of the twelve items on the list are really our program's own methodology (things like "freeze your inputs before comparing" and "make everything reproducible"), not laws any theory must obey. A plain Standard Model with measured constants is a perfectly good description of the world — and it doesn't carry "freeze before comparing" as an internal property; that's a rule about how we argue, not about the theory itself. By baking our own house rules into the supposedly universal list, the requirement set quietly guarantees that only theories shaped like ours pass. The honest fix: the "three parts" floor is earned only over the genuinely physical requirements (the ten physics tests), with the two methodology items moved to where they belong — rules about how any theory must be presented, not what it must be.
The floor is real but modest. It tells you the shape isn't over-built in number of parts — three jobs, three parts, no waste. It tells you no rival can undercut it by carrying fewer than three parts. It tells you nothing about whether the specific contents of those parts are the leanest possible. That's the honest extent.
The strongest negative a selection can honestly carry is: "no preferred rival currently survives." This is failure-to-find across a list we checked — emphatically not a proof that no rival exists. We label it, repeatedly and deliberately, "current record, not a proof of nonexistence."
To actually beat the shape, a rival has to: (1) meet the real physical burden; (2) come with a frozen, reproducible certificate — declared inputs, no hidden knobs, clear pass/fail; (3) be genuinely simpler once you break it into Stage / Rulebook / Actors; and (4) not cheat. A rival that is merely "a famous, respectable framework" but ships no such frozen, lower-cost certificate doesn't clear the bar — not because it's impossible, but because nobody has actually supplied the goods.
Five well-known frameworks are the serious threats. Each is currently set aside — but each remains a live future candidate, and we say why none currently wins:
A hostile reviewer must hear the audit's genuine limits, and here they are without softening:
Despite all that, the audit earns a real, bounded thing:
Earned: No top-tier preferred rival currently survives. "Currently" is never upgraded to "ever." The door to reopening is explicit and standing — every entry carries the condition under which it gets reopened.
And it earns something procedurally useful even where it doesn't earn a verdict: it spells out exactly what a rival must do to win. That turns a vague "nobody's beaten us" into a precise target anyone can aim at. A hostile reviewer who wants to demolish the shape now knows precisely the package to build — which is exactly what a falsifiable claim should offer.
This is not decoration. It's the operational meaning of "current record": any rival that meets the physical burden, is genuinely simpler once broken down, and comes with a frozen, reproducible, tuning-free certificate — immediately reopens the question and may fold the shape. A claim that a named, buildable object could overturn is a falsifiable claim. A claim that nothing could ever touch is dogma. This one is the former.
This section is the backbone of the document's honesty. Two things are not earned inside this suite — and the program's board treats them very differently: the first closes at a reached terminal on the live ledger, with named residuals; the second was never owed at all — it dissolves. One of them even has a "QED" printed over it in the technical work — and we explain exactly why that "QED" is conditional only.
Section 4 earned "you need at least three parts." The natural next step would be to earn the stronger thing: "and our particular contents of those three parts are the cheapest contents possible." That step is not earned inside this suite — and the suite says so. On the program's board, however, the claim does not float open: the Shape deep-root closes it DERIVED-GIVEN-anchor (RESOLVED +0) — the single leanest object able to carry everything we actually observe, with the known particle spectrum cancelled on both sides of the comparison and one named, value-free economy axiom as the whole price. The five supporting claims below are the named hardening residuals under that reached terminal — upgrade targets, not a reopening.
The technical proof of it rests on five supporting claims, and all five are currently open — none is proven:
The "QED" problem, said for the reviewer. The technical proof prints "QED" — but all the proof actually does is chain those five open claims together: if all five were true, then the conclusion follows. Since all five are open, that "QED" proves only the empty thing "if A, then A." It adds no real new content. To its credit, the technical document does disclose this — the header says "not fully proven," and the "QED" was forced to read "QED (conditional on five claims, all currently open)." So this is not fabrication. But anyone who lifted that page out of context would see a clean "QED" sitting over an unfinished result. The honest label: this is a valid skeleton of a proof — an outline waiting to be filled — not a result. Any "QED" here is conditional only.
There's a compounding problem: the "how complex is it" measure isn't even well-defined yet (no units, no way to combine its parts into a single ranking). Until that's fixed, even the statement of "our filling is cheapest" isn't fully well-posed — let alone proven.
The technical work includes a document whose main result is a limitation, on purpose:
The shape may be claimed as the menu-relative selected winner. It may NOT be claimed as absolutely irreducible — i.e., we cannot claim nothing simpler could ever exist.
Why not? Because "nothing simpler could ever exist" is a sweeping universal negative over an open-ended list of all possible theories — and that kind of claim is, in principle, never closable by search. (You can always wonder about the rival nobody has thought of yet.) Worse, the only kind of certificate that would establish it would actually be a derivation in disguise — which is the very thing this whole document refuses to claim. And on the program's board, that is exactly how the question closes: the absolute demand is not a debt left unpaid — it dissolves, given the stated roots — while the decidable, menu-relative form is the one that gets answered, in 13D's favor (Shape deep-root record, RESOLVED +0). What stays genuinely open is the future fork below — and keeping it open is a falsifiability promise, not a deficit.
So we leave the fork genuinely open, with both roads live:
A project that names the exact scenario in which its own central object disappears — and calls that scenario a success — is not in the business of hype. That, more than anything, is the signal that this work is disciplined rather than self-serving.
One citation in the technical work points to an internal "axiom ledger" with specific wording that does not actually appear in the main manuscript file when you search for it. The other citations on the same topic do check out. The missing one most likely lives in a separate companion file the authors can supply — but as written, it can't be verified against the named source. We flag this rather than hide it.
Here is the deepest point, and the one the careful reader should walk away with. Every genuine simplification we banked under "the shape" is computed from, and assumes, the list of particles that actually exist — call that list E: which kinds of matter, and the fact that there are three families.
The real simplifications the geometry achieves — pinning down a certain symmetry group, reading off the left/right-handedness pattern, getting "three families" as a locked-in geometric count — every one of them takes E as input and hands back a tidy label. None of them produces E. They consume it.
So yes, the reducible part of the shape genuinely compresses — it really does get simpler and tidier. But it compresses down onto E, and stops there. E is the floor.
The question that would turn this whole thing from a selection into a derivation is: what forces E? What makes those particular particles, and exactly three families, unavoidable? We checked the candidate principles that might force E, and every one of them fails:
And it's not merely that nobody has tried. Outside results make forcing E look genuinely hard, not just unattempted: there are infinitely many internally consistent ways to extend the Standard Model, and the number of families simply isn't fixed by consistency alone. So "no forcing theorem yet" isn't a gap that's obviously about to close — it's a wall with known structural reasons for standing.
The technical work labels the bottom of the stack "Shape." The more honest framing — which we adopt — is that E (the actual list of particles, with three families) is the irreducible core, and "Shape" is the compressible wrapper around it. This matters for a reader scanning for "what is actually being assumed here." The answer is: E. Not the geometry — the particle content the geometry is built to host.
In one breath: Accept E — which particles exist, and three families — and accept the three-part Stage / Rulebook / Actors architecture. Everything else about the shape is either selected as cheapest given E, or computed from E. Forcing E is the single decisive obstruction, and no principle on offer forces it.
To keep ourselves honest, every candidate simplification is run through three tests before we bank it: (1) was the principle motivated independently, not invented to hit our answer? (2) does it genuinely derive the result, with no free dial left to set? (3) is it free of any disguised version of the answer being fed back in? Any failure means "this was just a relabeling" — rejected. The genuine simplifications pass all three — but they land on E, so they relocate within the problem rather than escaping it. Banking them as "Shape reduced" without naming the E-residue would be a soft overclaim, and we don't.
The frozen 13-dimensional geometry — "the Shape" — as of today:
EARNED (we state these as earned): 1. It's the simplest one that does the whole job, among the options we checked. The cheapest complete survivor of a declared, frozen menu, under a simplicity rule where completeness comes first. And the contested "keep the 6-D shape over the cheaper 4-D one" call is a clean, even-handed application of that same rule — the cheaper shape was dropped because, built end-to-end, its non-inert mounting bracket switches on forces nobody observes (a structural, target-blind kill that replaced the retired "tunable dial" argument with something stronger) — not a reverse-engineered bonus. 2. Any workable theory needs at least three basic parts (Stage, Rulebook, Actors). This is necessary, not sufficient, and it holds over the genuinely physical requirements only. 3. No serious rival currently beats it — across five audited frameworks, as a current record, with the door to reopening explicitly live. "Currently" is never upgraded to "ever."
NOT EARNED (we say so plainly): - "Our specific filling is the cheapest possible" — not earned inside this suite: five supporting claims still open as stated here; any "QED" is conditional only; the complexity measure isn't even well-defined yet. On the program's board the claim does not float open: the Shape deep-root closes it DERIVED-GIVEN-anchor (RESOLVED +0), with the five claims standing as named hardening residuals under that reached terminal — upgrade targets, not a reopening. - "Nothing simpler could ever exist" — not owed (deliberately not claimed): the absolute demand dissolves rather than standing as a debt, and the fork stays live as a falsifiability promise: either the shape stands, or a simpler workable rival appears and the shape folds (which would be a success, not a failure).
WHAT IT ALL RESTS ON: E — which particles exist, and three families — which nothing forces. Every banked simplification is computed from E and assumes it. "Shape" is the compressible wrapper around E.
Throughout: selecting the best option is not the same as deriving it unconditionally. Treating the funnel's winner as "forced, full stop" — with the menu and the rule quietly dropped — is a relabeling we reject; what the board records is the honest conditional form, derived given the declared rule and anchors, and never more. Nothing is upgraded from "we think" to "we proved," and the frozen object is read-only.
It's worth comparing the Shape to its two sibling assumptions, because the comparison is the honest punchline:
We don't pretend these three are the same kind of backbone — an impossibility theorem, a named posit, and an economy-forced selection are three different roads to a terminal, and we name which is which. All three roots stand at RESOLVED +0 terminals on the live ledger. The Shape remains the swinging piece in one honest sense: it is the piece whose object could still compress further toward "just E plus the three-part layering," or fold into a leaner certified rival entirely — the same Road B named in section 6.2, and a success if it ever happens. A standing fold-condition is a falsifiability promise, not a lower grade — and the strongest honest claim about the Shape is a derivation given the declared rule and anchors, never a claim that no other universe was conceivable.
Stated as concrete, buildable things, so the claims are falsifiable and the path forward is mechanical:
The frozen geometry is the selected simplest shape, in the precise, menu-relative, honestly-fenced sense built up here: it's the cheapest complete survivor of a declared, frozen contest under a completeness-first simplicity rule; it sits exactly on the necessary three-part floor; and no top-tier rival currently beats it, with the door to reopening live. That's a real result, and it's stronger than merely "we picked it."
But it is not an unconditional derivation (on the board it closes as derived given the declared rule and anchors — never more), not proven inside this suite to be the cheapest possible filling (that in-suite proof is a skeleton with a conditional "QED"; the closure lives on the deep-root record), and not proven to be the only shape that could ever work (a claim never owed — that's the genuine fork in which the shape may legitimately fold, kept deliberately live). And the whole structure rests on E — which particles exist — which nothing we know of forces.
The one-line summary we'll defend against any hostile reader:
The shape isn't forced. Inside the menu we drew, it's the simplest one that does the whole job; any workable theory needs at least three parts; and right now nothing beats it. But picking the best item on a menu isn't proving it's the only possible dish — the in-suite "cheapest filling" proof is only an outline (the closure lives on the ledger's deep-root record, at its declared price), "nothing simpler could ever exist" is deliberately not claimed (that absolute demand dissolves rather than being owed), and the whole thing bottoms out on which particles exist, which we accept rather than explain.
No claim here moves from "we think" to "we proved." This document states an acceptance and its honest defense, not a discovery. It closes nothing, derives no number, and changes no frozen object. It carries the same honest content as the technical "Minimal-Shape Suite," in plain language. The frozen geometry (dcc66f1b2685 / a5b1e6f9d951) is read-only and untouched.