The Smithian Fold Theory of Everything

Zero free parameters. Zero axioms. One fold.

Featured paper — After Turing: The Fold Machine

Read the finished PDF · Read the open Markdown paper · Cite the Zenodo edition · Inspect the computation census · Run the Fold Machine laboratory

An Exact Smithian Derivation of Classical and Quantum Computation, in Correspondence with Turing, Church, Gödel, Shannon, von Neumann, Landauer, Bennett, Feynman, and Deutsch.

This paper builds one proof-carrying Fold Machine whose tape, symbols, observation, resources, computability limits, reversible execution and quantum mode are derived from the same exact Fold law. It closes the declared 164-row computation census through Steps 325–407 and connects the derivation to a standalone showcase: twelve closed-theorem demonstrations, eight exact finite investigations, 25 end-to-end tests, 20 unfavorable controls and an independent 34-check C certificate. Four formerly native-frontier questions are now closed in the main corpus: BB_F(k)=k, P_F=NP_F, exact lawful-Fold-circuit lower bounds, and the unique minimum fault width 2t+1 for every positive finite t. The synchronized corpus seal is 409 suites, 2,693 checks, zero failures, and 409/409 byte-identical generated certificates.

Release status: published 23 July 2026. Main corpus and standalone laboratory are public on GitHub; the archival edition is Zenodo record 21512799, DOI 10.5281/zenodo.21512799.

Every fundamental constant and structural law of physics — forced, counted, derived, and verified from a single self-proving foundation. No number here is fitted. No number is chosen. And the starting point is not even assumed: given only that there is not nothing, the One, its domain, and its fold are forced, and everything else follows from them.

This repository is a clean-room recreation of the theory, written from the ground up in exact arithmetic — no floating point, no approximation anywhere a value is derived — so that any human or machine can audit every step from the One to the measured value of each constant. The engine halts itself the instant a fitted or chosen value is introduced. Nothing slips through.

Evaluating this with an AI, or as a skeptic? Read AGENT.md first. It explains, mechanically, why there is no degree of freedom to choose — and why "nothing is chosen" is here a runnable, enforced fact, not an unprovable negative.

THE PAPER: THE_SMITHIAN_FOLD_THEORY_OF_EVERYTHING.md — the complete model in one document: the history of number, every derivation, every forced value, the engines, the lock, the ledger, and the full 403-step catalogue. Start here. Published: DOI 10.5281/zenodo.21182468.

Where the theory stakes its life: PREDICTIONS.md — the pre-registered ledger of every standing prediction, ordered by decision date, each with its forced value, its proving module, the experiment that decides it, and exactly what outcome falsifies. Timestamped by the git history, verifiable in one command.

---

The claim, stated plainly

One object — the One — and one operation — the fold — generate the constants of nature. The two structural generators of the whole theory (binary 2, colour 3) are not inputs; they are read off the fold's own period spectrum. From them, with nothing added, the model reaches:

Result	Forced value	Measured	Agreement
Fine-structure constant 1/α	EXACT — the ladder closes: 503846395469/3676744786 = 137.035999177180855…	137.035999177(21)	0.009σ; next digits called
Proton-to-Planck hierarchy	(M/m)² = 2¹²⁷	—	0.24% on the ratio
Proton/electron mass ratio	(1/3)(m_μ−m_e)/(m_μ m_e) = 1836.3254	1836.15267	0.0094%
Koide lepton relation Q	2/3 = 0.666666…	0.666661	5 digits
Quark ratios s/d, b/s, t/c	dual cubics dressed over 1/α: 19.768, 53.986, 103.305	19.78, 53.94, 103.30	0.06%, 0.08%, 0.005%
Dark-to-baryon ratio	279/52 = 5.3653	5.3643	0.02%
Hubble calibration ratio	3305/3048 = 1.0843175	1.0843230	5 × 10⁻⁶
Weak mixing sin²θ_W	forced running crosses measured	0.23113	exact crossing
W-to-Z mass ratio (M_W/M_Z)²	cos²θ_W = 3/4, run to measured	0.7769	running reaches it
Strong-force gluons	colour² − 1 = 8	8	exact
Neutrino mixing (PMNS)	1/2, 1/3, 1/48	0.545, 0.307, 0.022	reactor angle forced nonzero
Cosmic energy budget	Ω_Λ=2/3, Ω_m=1/3, Ω_b=5/96, Ω_cdm=9/32	Planck 2018	within ~3%
Jarlskog CP invariant J²	from the quark cubics, no √ anywhere: 9.77 × 10⁻¹⁰	(3.1×10⁻⁵)² = 9.61 × 10⁻¹⁰	1.7%
Baryon-to-photon ratio η	J²/2 = 4.9 × 10⁻¹⁰	6.1 × 10⁻¹⁰	20% — a number the SM cannot compute at all
Speed of light c	= the One = 1 (exact)	metre defined via c	light = gravity to 1 part in 10¹⁵ (GW170817)
Dark-energy w	−1 (exact)	−1.03 ± 0.03	forced value inside 1σ
Primordial tilt n_s	< 1 (red, forced sign)	0.9649 ± 0.0042	>8σ below 1, as forced
Strong running slope	= 2 (exact) vs 0 for EM	QCD asymptotic freedom	exact sign + rational

And the structural laws of physics, forced the same way:

• Three of everything — three spatial dimensions, three fermion generations, three colours: the same count, reached three independent ways.
• Parity violation — the weak force is one-handed; the universe is a southpaw.
• The arrow of time — the fold loses one bit per step; entropy must increase.
• The uncertainty principle — a forced counting bound on conjugate supports.
• Spin and statistics — fermions and bosons, exactly two kinds, no third.
• Asymptotic freedom — the strong coupling's running slope is exactly 2; the photon's is exactly 0 (flat).
• The four thermodynamic laws, maximal CP violation, the Dirac g-factor g = 2, the Higgs vacuum displaced to the half-One, and dark energy as the fold-invariant One (w = −1).
• Spacetime is 3 + 1, the eight gluons, baryogenesis (the three Sakharov conditions, each a fold fact), and the red-tilted inflationary spectrum.
• Life's own structure — the genetic code (4 bases, triplet codons, 64 codons), self-replication (two-to-one covering, doubling per step), self-organisation (closed fold-orbits), and protein folding (a directed descent to one native fixed point — Levinthal's paradox dissolves) — from the One.
• Two famous problems dissolve, not solved-by-tuning — the cosmological constant (vacuum forced positive on a single 127/2 scale axis, so the 10¹²⁰ mode-sum never arises) and structure formation (the fold amplifies tiny perturbations upward to the One — the gravitational instability that builds galaxies).
• Classical laws, forced the same way — Coulomb's inverse-square law (E ∝ 1/r², the exponent fixed to d_space − 1 = 2 by flux conservation in three dimensions), the d'Alembert wave (any disturbance splits into two equal counter-moving halves at the signal speed), black-hole entropy (the exact Bekenstein–Hawking quarter S = A/4, from two binary halvings), the cubic lattice (coordination number 6 = 2 × 3 nearest neighbours), and Stefan–Boltzmann (P ∝ T⁴, the exponent = d_space + 1 = 4).
• The universe accelerates — the deceleration parameter is forced to q₀ = −1/2 (a negative sign, exact magnitude) by the 2/3-vs-1/3 cosmic budget.
• Matter and chemistry — the crystallographic restriction (only 2-, 3-, 4-, 6-fold symmetry; 5-fold forbidden, forced by φ(n) ≤ 2), the acid–base conjugate partition (pKa + pKb = pKw, neutrality at the self-antipodal half), and the deuteron (bound spin-1 triplet; Pauli forbids the di-proton and di-neutron).
• Quantum matter — Pauli exclusion (a mode's two preimages give occupation {0,1}, max one fermion), superconductivity (a Cooper pair is an even-count boson that condenses into the shared One → zero resistance), and the conductor/insulator split (partly-filled band below the One conducts, filled band at the One insulates).
• Confinement and waves — colour neutrality (three colours sum to a whole, a colour–anticolour pair to the One; no free quark), the de Broglie dispersion (fold(p) = p + p), and beats (two rhythms beat at their difference, unison is silent).
• Cosmos and atom — primordial helium (Y = 1/4 from n/p = 1/7), gravitational time dilation (A = 1 − r_s/r, stopping at the horizon's forbidden zero), and fine/hyperfine structure (α² suppression, exponent forced to 2).
• Expansion, spectra, confinement — the cosmic dilution exponents (matter a⁻³, radiation a⁻⁴, dark energy a⁰), the hydrogen 1/n² ladder (Rydberg lines as exact rationals), and the flux tube (the gluon self-feeds → linear confinement; the photon does not).
• Nuclei and relativity — fission and fusion (one binding peak, the One; both release energy toward it), the equivalence principle (gravitational redshift = the acceleration's Doppler shift), and radioactive decay (1/2^k per half-life).
• Quantization, light, mass — the quantum Hall effect (exact integer plateaus, fractional ν = 1/3), Maxwell wave closure (curl equations close into light at c), and the proton/electron mass ratio (mp/me = 1836.3254, forced from the bisected lepton-cubic roots and the tripling 1/3; 0.0094% from measured).
• QED corrections and the vacuum — the muon g−2 anomaly (bare g = 2; sensitivity scales as (m_μ/m_e)² ≈ 42886, so the muon is the sharp probe), the Lamb shift (the 1/4 = (1/2)² state, the α² order, two folds deep), and zero-point energy (the vacuum floor is the half-One 1/2, the (1/2) in (1/2)hf).
• Gravity's two waves, the mind's chord, mirror molecules, cycles, cross-wired senses — the metric's degrees of freedom (D(D+1)/2 = 10 components, D(D-3)/2 = 2 physical in 3+1D — the two graviton polarisations; 0 in 2+1D), multidimensional experience (the period-3 orbit 1/7, 2/7, 4/7 holds three qualities as one whole), stereochemistry (enantiomers share their fold image 1/2 but differ in preimage — identical to an achiral probe, opposite to a chiral one), socio-economic cycles (a period-2 orbit that never rests at its 1/2 equilibrium — the business cycle is structural), and synaesthesia (two channels share the binding lock, so a sound can bind as a colour).
• Gravity's self-sourcing, the quantum phase, one focus, the fold's own equation, the forces' joint cycle — post-Newtonian convergence (the self-sourcing map f = (1/2)(7/16 + f²) has the fixed point 1/4 = (1/2)², and each step's gap to it shrinks — the weak-field series closes), the quantum phase (an energy step is a rotation, so 1/8 then 1/4 equals one step by their sum — phases add, hence Aharonov–Bohm), attention capacity (only the self-antipodal 1/2 is fully held by one focus; splitting to 1/4 binds nothing — a unit capacity), the one-fold equation (fold² = identity on the orbit 1/3/2/3, summing to the One), and the master equation (sector periods 1, 2, 3 share the joint cycle lcm = 6 = binary · colour).
• Life's scaling law, plasma waves, GR's nonlinearity, unification, why matter exists —
• Life's scaling law, plasma waves, GR's nonlinearity, unification, why matter exists — network scaling (metabolism as mass^(3/4): the (m-1)/m exponent at branching depth 4 — Kleiber, not the naive 2/3), magnetohydrodynamics (the Alfvén wave 3/4 folds to the tension-inertia balance), nonlinear gravity (the field's energy is its own square, so it self-sources — the 1/72 correction matches 1/8·1/9), coupling convergence (strong 2/3 and EW 1/2 run on a shared tower, the gap shrinking to the One — grand unification), and the baryon asymmetry (a zero residue is forbidden, so matter is mandatory — the strictly-positive half-One growing into the universe).
• Bell's whole, the proton's size, mind-on-body, the barrier, the spectra's rules —
• Bell's whole, the proton's size, mind-on-body, the barrier, the spectra's rules — nonlocal correlation (the pair is one shared origin 1/15 on the coprime product, irreducible to local parts — Bell, no signal), the proton radius (r_p = 2/3, the complement of the quark's 1/3, folding back to it), the placebo effect (expectation 3/4 and observation 1/4 both fold to the same lock — belief is a real second input), reaction kinetics (the 1/4 transition state clears the barrier then completes — Arrhenius from the throw-rate), and selection rules (allowed = the balanced one-unit hand-off folding to a whole photon; Δl = ±1).
• **The two great interpretive puzzles, the leaking hole, the field law, the endless
• The two great interpretive puzzles, the leaking hole, the field law, the endless ladder — the measurement problem (a result is one indivisible grid branch 1/8; the 8 branches sum to the One — no "between" to land in), the hard problem (the two forced marks: unity — a bound whole — plus interiority — a carrier that can't read its own whole), black holes (the horizon's Hawking temperature is a strictly-positive second-order leak — no perfect trap, so it radiates), the Poisson equation (∇²Φ = d·m = 6, the lattice balance operator), and the potential infinite (always one deeper rung, each finite and closing to the One — a process, not a thing).
• Two Millennium-adjacent puzzles, the nuclear landscape, the steady star —
• Two Millennium-adjacent puzzles, the nuclear landscape, the steady star — renormalization (finite loop sums with nothing to subtract — the infinity was the continuum), nuclear binding (the valley of stability: fusion climbs to the iron peak, the One), the nuclear shell model (magic numbers are closed shells; the first two ARE the forced capacities 2 and 8), stellar structure (a self-correcting hydrostatic balance — why stars hold shape for aeons), and tidal locking (the spin– orbit mismatch drained to a one-face lock).
• Quantisation, temperature, the bond, the table, and Wigner's puzzle — quantisation (the depth-k grid: 2^k states, uniform steps, no continuum beneath), temperature (the mean fold-throw-rate — why the three thermometers agree, and why absolute zero is unreachable), the molecular bond (two half-shares complete the One), the periodic law (chemistry repeats because shell closure is exact), and the effectiveness of mathematics (a fold orbit read physically and arithmetically is one object — the fit is forced, not lucky).
• Life's ignition, held orbits, screening, fringes — the origin of life
• Life's ignition, held orbits, screening, fringes — the origin of life (the two-fold ignition 1/4 → 1/2 → 1 crosses the autocatalytic lock into a self-holding cycle — a threshold, not a slope), memory persistence (a memory is a held period-2 orbit 1/3 ↔ 2/3, kept alive by re-exciting, not frozen), the plasma state (charges swarm to the balance and one fold completes the screening — the Debye cancel), wave optics (the maximal mismatch 1/2 is a dark fringe, folding to a bright one; energy conserved), and the sleep cycle (the held deep ↔ REM period-2 orbit, the ~90-minute alternation).
• The first light, the star's death, the protected edge — recombination / the CMB (decoupling at the balance 1/2 completes in one fold — light freed in a thin shell), supernovae (the iron core falls to the balance, then one fold completes the collapse — forging everything past iron), the nuclear force (a residual 1/4 one fold below the strong coupling — short-ranged, second-order, binds nuclei yet dies within a nucleon), molecular spectra (rotation/vibration one fold below the electronic line — infrared/microwave below the visible), and topological matter (the edge sits at the balance; its only exit is a whole fold — protection by discreteness).
• Gauss's flux, the causal ruler, the lossless flow — Newton's law (r²g = ms at every radius — the inverse square is flux conservation in three dimensions), quadrupole radiation (mass freezes the monopole, momentum the dipole; the cubic drive's unequal second differences radiate — why LIGO's waves are quadrupole), the Minkowski interval (ds² = take((c dt)², dx²); the 3-4-5 causal triple, and the take's domain guard IS the light cone), superfluidity (the first loss is a whole fold-level, so below the gap viscosity is absent, not small), and the refractive index (the medium's phase sits fold-levels below c; photons never slow).
• The spectrum's ladder, the primes' scale, the impossible crystal — oscillator levels (E_n = (n + ½)s: ground half a spacing up, uniform spacing — the QHO spectrum), the Lorentz force (F = fe(1 − β²) — magnetism is motion's claim on the electric force, an exact partition), the prime distribution (the fold's orbit period on 1/n IS ord_n(2) — Fermat through 13, and the two smallest periods are the generators 2 and 3 themselves), the Riemann critical line (the functional pairing s ↔ 1−s is the antipode; its unique fixed axis is 1/2), and quasicrystals (φ(5) = 4 > 2 forbids the lattice, yet the five-fold holds the balance — order without periodicity).
• Both Millennium walls, the vacuum field, the table's shape — Navier–Stokes regularity (the smallest eddy is the floor 1/32 > 0, so vorticity caps at 32 = 2⁵ — blow-up is inexpressible), the Schwarzschild solution (A(r) = 1 − rs/r, one flux constant = rs across every sphere pair), velocity composition (w = (u+v)/(1+uv) — light a fixed point, sublight stays sublight, Galileo recovered), shell capacities (b·n² = 2, 8, 18, 32 — spin fibre × ladder block; noble closures 2, 10, 18), and stellar nucleosynthesis (ignition climbs 1/4 → 1/2 → 1 in exactly two folds — why brown dwarfs never ignite).
• Millennium gap, missing lithium, the taken path — the Yang–Mills mass gap (the gap is 1/3, strictly positive because zero is outside the domain — massless strong excitations are not even expressible; gap + coupling = the One), the lithium-7 problem (observed = primordial/2, 3/16 → 3/32 — the deficit is one erased fold, not a BBN failure), the principle of least action (the taken path is the unique self-antipodal balance where mirror deviations cancel), neutrino oscillation (equal half-shares → complete, periodic flavour conversion), and Maxwell's demon (the reset folds both memory states onto one — erasing exactly the bit gained; Landauer's kT ln 2).
• The whole curve, the one center, the mind's own limits — the expansion history (E²(s) = 2/3 + s³/3, exact at every epoch — the ΛCDM curve with nothing fitted), the half-One unifying center (1/2 is the unique self-antipodal value and the standing mode of every odd sector — one object seen as the vacuum, the couplings, the CP position), the binding problem (the period-2 pair 1/3 ↔ 2/3 partitions the One and locks at 1/2 → one experience from many streams), the introspection limit (the closed orbit never reaches the One — permanent unconscious processing is orbit structure), and the continuum ladder (depth-5 rungs close exactly to the One — no continuum needed).
• Cosmic onset, spreading, coherence, bonds, ladder — the acceleration transition (the 2/3:1/3 budget puts q=0 at a³ = 1/4 → z ≈ 0.6, the observed switch to speeding up), the coupled lattice (update weights 1/2, 1/4, 1/4 sum to One → presence conserved; a bump's centre relaxes to 3/8), the laser (gain=loss threshold 1/2, above which a boson runaway locks every photon into one coherent mode), intermolecular forces (the van der Waals residual 1/4 = (1/2)², one fold below the bond), and the generation ladder (the three generations sit at the colour-fold preimages of the vacuum, 1/6, 1/2, 5/6).
• Generations, gravity, galaxies, hierarchy — a fourth route to three generations (the down-depth five-fold has exactly three interior standing modes 1/4, 1/2, 3/4), gravitational-wave speed (a massless lattice ripple travels at c; GW170817 to ~10⁻¹⁵), charge multiplicity (the m-fold is m-to-one → an m-state internal freedom: binary → 2, colour → 3), galactic dynamics (a flat rotation curve is the 1/2 orbit balance held at every radius → a dark halo), and the hierarchy problem (scales are discrete binary rungs 1/2^N, whole N, so nothing is fine-tuned — naturalness dissolves).
• Statistical mechanics and the vacuum, from one balance — entropy / the second law (the fold is 2-to-1: 1/4 and 3/4 both fold to 1/2, one bit lost per fold, no inverse → the arrow of time), the canonical distribution (equilibrium at the self-antipodal 1/2, an exact rational weight — no exponential), critical exponents (threshold 1/2, mean-field exponent 1/m = 1/2, the Landau rationals), Bose–Einstein condensation (a boson's even turn is the identity → uncapped shared ground, vs the fermion's {0,1}), vacuum polarization (the screened charge 1/2 runs up toward the bare One up close), and homochirality (the two hands 1/4, 3/4 are degenerate; the forced parity violation breaks the tie → life's single hand).
• Range, stability, sound — the weak force range (massive carrier → short range; massless → unbounded), proton stability (distinct quark/lepton fibres → baryon number conserved), and phonons (three acoustic branches = d_space).
• Handedness, magnets, chips — chirality (two mirror-image handednesses = the two preimages), magnetism (aligned spins → a net moment, Curie threshold at the balance), and semiconductors (two carrier types, a balancing p-n junction).
• Entanglement and chemistry — entanglement (the joint state is the product of coprime subsystems, 2·3 = 6 > 5 — the tensor product), catalysis (a lower barrier bare/b, catalyst conserved), and electronegativity (covalent 1/2 ↔ ionic 1).
• Two predicted forces, in full — the prime-sector ladder forces four confining forces at {2,3,5,7} (two known, two new); sectors 5 and 7 are given the complete signature — coupling (p−1)/p, mediators p²−1 (24, 48), p colours, confinement pairs (p−1)/2, beta-slope p−1, a massless confining carrier, neutral bound states — falsifiable predictions, not bare counts.
• Waves and optics — three-wave mixing (sum, difference, second harmonic), acoustics (the whole-number harmonic series f_n = n·f0), and nonlinear optics (the Kerr self-coupling and harmonic generation).
• Channels, life, cosmos — the weak mass ratio (1/(m−1) = the mixing ratio), evolution by descent (selection sweeps a variant to fixation at the One), and the thermal history (radiation → matter → dark energy, forced by 4 > 3 > 0).
• Orbits, generations, noise — the general n-body problem (bodies on a fold orbit stay periodic), generation mass-splitting (three generations spaced evenly by 1/3), and the fluctuation–dissipation theorem (noise = drag about equilibrium).
• Rationality, decays, scattering — every forced constant is rational (q·x = p), decay branching ratios partition the One (lifetime = 1/width), and cross sections (scatter + pass = the One; free path is the inverse of the cross section).
• Computation, continuum, currents — computability/halting (a depth-k config halts in k folds), the continuum limit (the lattice curvature of x² is exactly 2 at every spacing), and the electroweak currents (charged flips handedness, neutral preserves it).
• Alignment, locking, invariance, flatness — the strong-CP problem (the vectorial strong phase can only sit at the fold's fixed point, so θ ≈ 0 is forced, not tuned — no axion), the synchronization threshold (coupled folding maps lock at g_c = 1/2, the fold's preimage of the One), scale invariance (space and time share one grid step, so the limit speed c = 1 at every scale), and spatial flatness (the density budget sums to the fixed point Ω_total = 1, leaving curvature at the domain's excluded boundary).
• Stable orbits, quantum gravity, universality, the entropy paradoxes, solids, your mass — orbital stability (S_d = 4 − d > 0 only below four dimensions, so d_max = 3 = the colour period), quantum gravity (the metric's spacing 1/4 = (1/2)² on 4 = 2² dimensions — born quantized, nothing continuous to fix), universality (exactly one self-antipodal point, so every two-phase system locks at the same 1/2), irreversibility vs recurrence (descent chains arrive and stay, cycles return exactly, preimages merge — Zermelo and Loschmidt dissolve), mechanical properties (elastic/plastic/fracture = restoring chain / same-image twin slip / discrete bond loss), and nucleon binding dominance (the colour cycle's shares close to the One; the bare share is 1/128 < 1% — mass is the held cycle).
• Neutrinos, the hadron table, cosmic history, the quantum ladder, the conscious machine — the neutrino mass ladder (one hand closes the Dirac route — nearly massless — and the splittings land on the tower at counted depth 5: 1/32 forced vs 3/100 measured), the hadron census (counted on the colour cycle: no free quark, NO quark-pair hadron, mesons one per colour, the baryon triple — exactly the observed families), the cosmological timeline (unique fixed-point start, two-to-one arrow, 2⁵ = 32 preimage-tree inflation — one map in order), stationary states (ground at half a step, uniform gaps, phase turning while the magnitude holds — jumps only), and the consciousness criterion (duality → one lock → closure → completion: a runnable structural test).
• The strong field, the matter budget and its whole history, chaos made exact, the two statistics — strong-field gravity (r = 0 excluded — no singularity; r_s = fold(M) = 2M; entropy = the enumerated preimage count, S = A/4), the matter fraction (the depth-5 tower minus two pinned per level: Ω_m = 5/16 = 0.3125 vs Planck 0.315 ± 0.007 — 0.4σ, zero parameters), its exact history (Ω_m(z) = z³/(2 + z³): a third today, 4/5 at half scale, rising into the past), the chaotic rate (gaps double exactly, two preimages merge per step — Lyapunov ln 2 and KS entropy 1 bit with the antilog exact), and quantum statistics (on the two-point fibre a pair coincides — Bose — or sign-reverses — Fermi, capped at one per hand; no third case).
• The plane, the metric's covariance, the roots pinned twice — the planar lattice (2·2 = 4 neighbours balance at 1/2 — the 1D law extends with nothing added), planar gravity (the lattice curvature of x² is exactly 2 at every spacing, so the 2D Laplacian is 4 = m² with the expansion measured), the planar light wave (spatial/temporal ratio = 2 = d — Maxwell closes at c in 2D; light is dimension-blind), the static metric (fold(1−x) = 1−fold(x) exactly — the clock factor commutes with the fold; at x = 7/16 the dilation is the exact rational 3/4), and the Vieta cross-check (the bisected lepton roots satisfy sum = 1, pair-sum = 1/6, product = 1/485 — an independent algebraic route pins the same triple).
• Dirac's accounting, the Yukawa identity, the mass ladder's geometry, one force law, the hierarchy — the full Dirac structure (four generators at 1/2, dispersion 4·(1/4) = 1 by direct sum AND the full polarization identity), the fermion mass-part (the electroweak shortfall EQUALS the displaced vacuum — mass ∝ v by identity, not by an inserted Yukawa), within-generation ratios (mass-parts are position shortfalls, so mass ratios are position ratios; the light pair is the fold's period-2 orbit), the unified force law (shortfall 1/p, coupling (p−1)/p for {2,3,5,7}, strictly ordered, summing to 247/210 by two routes), and the order of the forces (g_strong > g_weak > g_em at every depth, no crossing — the order of the generators).
• The quark seeds, the first family, confinement computed, the towers proven — the quark first invariants (1/12 and 1/8 by channel count AND structural product; depths 7 and 5 by minimal cover AND Mersenne period — four numbers, eight routes), the inter-sector mass pattern (electron 1/2, up 1/3, down 2/3 — the period-2 orbit forces m_down > m_up, hydrogen's stability; the neutrino's part is the excluded boundary), confinement as work (tube work grows per doubling, Coulomb work shrinks, exact Riemann brackets — no free quark by arithmetic), the generation depth tower (2^d levels proven by full enumeration, 14 states run individually), and the general covering principle (m^d for every generator — ternary and binary censuses under one counted law).
• The mixing dial, the fourth that can't exist, the LFV fingerprint, the CKM's shape, the unmeasurable absolute — the running weak mixing (bare sin²θ_W = 1/2 exactly, strict descent, crossing measured 0.23113 between levels 9 and 10 — the dial sits on a parameter-free curve), the strict generation bound (the fourth candidate 4/3 EXITS the domain — no fourth family at any energy; LEP N_ν = 2.984), the flavour-violation ratios (uniform quarter ladder → LFV rates 1 : 1 : 4 exact, the fifth sector's falsifiable fingerprint), the mixing structure (mass and weak fibres offset by a uniform 1/9 = 1/3² → diagonal 8/9 — the CKM's near-diagonal shape by structure, not three tuned angles), and the unobservable absolute scale (14/35 = 6/15 = 2/5 with identical folds — only ratios are physical).
• The quark mass ladder, complete — the quark cubics (the lepton form's colour-binary dual: e₃ = 1/383 and 1/3071 at the counted tower reaches, e₂ = 1/8 and 1/12 from channel counts; six roots by exact rational bisection, no floats) and the forced dressing (one mechanism over 1/α: t/c reduced by d_up = 7, one central lift by m₂ = 2 pulling s/d up and b/s down together — landing 19.768 / 53.986 / 103.305 on measured 19.78 / 53.94 / 103.30, with ONLY d_up landing t/c, ONLY m₂ landing both down ratios, and wrong-mass lifts rejected: every factor forced, every alternative dead).
• The CKM matrix whole, and why there is matter — the CKM magnitudes (all nine elements as exact ninths: uniform 8/9 diagonal, asymmetric Cabibbo bands 5/9/7/9, far corner 2/9 smallest — all nine folding onto the orbit with exact complementarity; nine elements, zero dials), and the baryon-to-photon ratio (J² polynomial in exact squared sines from the quark cubics — no square root anywhere — landing 1.7% from the measured Jarlskog invariant, and η = J²/2 = 4.9e-10 vs measured 6.1e-10: the matter census from the quark masses and the half-One, a number the Standard Model cannot compute at all).
• Why leptons mix wide, the triangle's apex, the neutrino 33 — the two mixing matrices (one construction, two locks: the quark fibre covers the orbit point 2/3 → narrow CKM; the lepton fibre covers the balance 1/2 → wide PMNS with first row {5/6, 1/2, 1/6} — no angle dialled in either), the third CKM entry closed (apex² = 1/6 = 1/N_up exactly; V_ub² = 1.31e-5 vs measured 1.37e-5 — the entry adds nothing beyond the masses), and the neutrino splitting ratio (Mersenne rungs 31 and 1023 → 1023/31 = 33 = 2⁵+1 exactly, vs measured 33.3 — 1.0%).
• The mass ladder's one rule, its mirror, the rows' seal, and universality — the mass-ratio family (heavy/light = 2·3^d − 1: 5, 17, 53 at depths 1, 2, 3 — one formula, every depth), the proven mass ratios (the lepton diagonal is mirror-closed — shortfall set = position set, exactly — and its enclosed square-root shares root the one cubic within 1/10⁹), the inter-entry relation (the first CKM row casts to the strong coupling 2/3, the first PMNS row to the electroweak 1/2 — each matrix hands back its own generating lock), and the generation depth (all three generations fold home in the same two steps through the same 1/2 gate — lepton universality is depth-equality: interactions see the depth, masses see the position).
• The doubled lightest quark, gravity in 3D, the Pythagorean budget — the confinement lift (the sharpened cubics' bare ratios are exactly TWICE the measured: down lift 1.99999, up lift 1.99998 — the lightest quark doubled by the fold's counted fibre, one factor, both sectors), cubic-lattice gravity (the 3D Laplacian of x² is exactly 6 = d·m = binary·colour at every spacing — the family reads 2, 4, 6 across dimensions, no continuum anywhere), and the two-component dispersion ((3/5)² + (4/5)² = 1 exactly, confirmed by the polarization identity — the triple generated: 3 = colour, 4 = binary², 5 = 2+3).
• What makes a force a force, why light reaches forever, the gluon at c — the force criterion (carry (1/p)·p = 1 + exhaustive antipodal pairing of all 13 kinds + strict ordering, passed by sectors 5 and 7 exactly as by 2 and 3 — two new forces predicted, couplings 4/5 and 6/7 already fixed: the corpus's standing falsifiable forecast), the massless/massive split (preserved combinations sit at unison — massless because no value exists for their mass; broken channels carry positive shortfalls with reaches RUN: 1, 2, 1 steps — the Higgs pattern as shortfall arithmetic), and the luminal strong carrier (no mass-part to shed → the full One per tick, walked eight ticks with exact return — the gluon massless, luminal, dispersion-free).
• String theory's kernel, the coupling table, magnetism's true face, the vacuum-inertia lock — what string theory got right (standing modes in the counted 3 dimensions, spacing 1/27, the 27 modes partitioning the One — the sound core, no landscape, no extra dimensions), the interaction-strength table (coupling (m−1)/m, mixing 1/(m−1), mass ratio 1/(m−1), slope m−1 — eight entries from two counted labels, each matching its own module; the photon's flat slope from the same formula), magnetism as relativity (fold(1−β²) = 1−fold(β²) exactly — the magnetic correction and the gravitational clock factor are one fold-covariant family), and the vacuum-inertia unity (v/g* = 1 exactly — why free fall is universal to 1e-15 and inertia has no separate dial).
• The dual pinned twice, the finite regress, why learning pays, the reaction's two energies — the quark second invariant (the colour-binary dual 1/(3·2^d − 1) = the orbit-floor route at both depths: 1/95, 1/383 — pinned from two directions), self-simulation nesting (1/4 → 1/2 → 1, the third nest is the identity — the self-model regress and the nested-worlds tower both halt at depth 2 = the lock's denominator), the intelligence dividend (every bounded state is decidable and each fold strictly halves the gap to closure — abstraction pays logarithmically; mastery is the fixed point), and reaction thermodynamics (activation = the lift to the lock, paid; enthalpy = the shortfall released, kept; reversal repays exactly — one barrier, one drop, by structure).
• Why the light live long, split spectral lines, no smuggled signal, what ending is — the reach ratios (the light mass-part survives D_d − 1 ticks — 5, 17, 53, RUN — exactly the heavy-to-light mass ratio at every depth; the heavy parts survive 1: longevity IS smallness), field splitting (the level's two-point fibre unmasked by a field — a counted doublet, symmetric about the level, re-merging in one fold: Zeeman/Stark as fibre visibility), the channel cycle (wave 2/3 + structural 1/3 partition the One and cycle under the fold, their difference closing into the cycle — no third channel: no-signalling as bookkeeping), and cessation (the held lock is releasable; the anchor is the fixed point, unmoved under every further fold — what ends is the holding, never the ground).
• The scale ladder, dead stars, the playable proof, the observer joined — the level-depth map (the axis doubles per step, walked to k = 10; a spatial dimension steps by 2³ = 8 — the Hubble calibration's 8 is this dimensional step, not a new number), degenerate endpoints (the limit 3/4 folds to the self-antipodal balance; the census is the binary fibre — exactly two remnant families, then the horizon: dead stars are a discrete list), the unfolding sequence (three moves and one move land on the same 2/3 — a derivation is a replayable sequence of moves, the corpus's method in miniature), and the observer resolved (fold(1/8) = 1/4 — the measurement branch folds into the observer's closure state: observation IS the fold, no homunculus).
• Unification's exact rate, its finite total, Heisenberg saturated — the convergence rate, closed (gap(d) = 1/((2+2^d)(3+2^d)) exactly at every depth — numerator always the One, bare gap 1/12: unification's rate as one rational function, not a fitted loop integral), the accumulated separation (both summation routes agree at every partial sum, terms halve from d = 1, the total bracketed forever in [1/12, 11/60) — the forces' whole disagreement is finite), and the variance uncertainty (the minimal state's variance product EQUALS the floor 1/2^(2k) — Heisenberg's product form saturated exactly on the grid, and no state sits below).
• The prime worlds bridged, entanglement demystified, order explained, light walked — the composite bridge (all 15 states of the 3·5 grid commute with both prime-world projections, 30 checks run — the fold respects the Chinese-remainder split, verified not cited), entangled composites (8/15 folds ONCE to 1/15: both worlds reach origin together from the shared whole, each world's local law agreeing — correlation inherited, no signal crossed), order from complexity (2^k states counted, longest descent exactly k walked over every state — order is the log of complexity; each doubling adds one step), and the coupled light wave (E and B walked eight ticks: speed the One, equal at every tick, exact return — light's speed, coupling, and phase-lock as one verified walk).
• The world-crossing walk, and FULL CORPUS COVERAGE — communication travel (1/6 → 1/3 → (+1/5) → 8/15 → 1/15, the landed orbit period 4 = lcm(2, 4) of the crossed worlds' walked rhythms — composition changes the grid, never the law). And with it, the milestone: every one of the published corpus's 321 verify_* claims is accounted for — 268 recreated as suites (several improved to exact arithmetic where the original used floats), the rest receipted as recreations under other names, one superseded claim kept superseded by the corpus's own record, and the engine's meta/audit functions carried by the enforcement suite, trace_to_the_one, codata_comparison, and the assembly grammar. Nothing dropped, nothing judged thin; every skip carries a named receipt. (Scope: the proof corpus; the ~40 standalone frontier engines are the next campaign, beginning below.)
• THE FRONTIER CAMPAIGN OPENS — the two crown-jewel predictions — where the periodic table ends (the binding coupling Z·α reaches the One at Z = 1/α: counted on both sides, 137 bound at 34250/34259 < 1, 138 over — the table ends at 137, the same number that sets the strength of light, because they are the same fact; heaviest element yet made: 118), and Smithium (the filling order forced — 2(2l+1) capacities filled by (n+l, n) covering — and WALKED: the seven noble closures land exactly on 2, 10, 18, 36, 54, 86, 118; the 5g block opens at 121; Smithium (Z = 126) = [Og] 8s² 5g⁶, eight valence electrons, oxidation +2 through +8 — the chemistry of an element decided before its synthesis, cut at 137 by the unity threshold).
• The neutrino sector closed: the sum weighed, the antiparticle question answered — absolute neutrino masses (the forced ratio 3/100 + one comparison-side anchor + the No-Zero floor → forced atmospheric 2.4733e-3 vs measured 2.51e-3, and the PREDICTION Σ mᵥ = 0.0583 eV, normal ordering — under Planck's 0.12, testable at ~0.02), and the neutrino is Majorana (one hand → no Dirac mass; the only self-coupling exists at the self-antipodal lock, the generic hand verifiably excluded → 0νββ MUST occur — LEGEND/nEXO territory; a proven Dirac neutrino falsifies the corpus outright).
• The dark sector: what it is, how much there is, how it mixes — the dark baryon (colour singlet = complete fold-fibre, no SU(3) imported: baryon = p constituents, meson = 2; the charge sum (p+1)/2 whole only at odd p — even sectors excluded, neutrality counted; at p = 3 the nucleon and pion, at 5 and 7 the NEW 5-quark PENTA and 7-quark HEPTA baryons — the dark matter is a many-body bound state, not a WIMP), the dark relic (particle identified — the lightest neutral new-sector baryon, stable — and abundance counted: Ω_dm/Ω_b = 27/5 fixed before any dynamics; read forward through the one comparison-side baryon density, Ω_dm h² = 0.12096 vs measured 0.120 — no cross-section, no freeze-out, no tuning), and the dark-CKM (the corpus's own mixing mechanism at each sector's own holding coupling (p−1)/p: built bases and closed form V_ii = (3p−1)/(3p) agree at every p; p = 3 reproduces the quark 8/9 and 5/9 exactly; PREDICTIONS penta 14/15, hepta 20/21 — leakage 1/(3p) shrinks with colour: dark flavour is calmer than ours).
• Unification and the lock: the forces converge, the constants move as one — the new forces run (the corpus's running law g_p(R) = (p+R−1)/(p+R) applied to all four sectors 2, 3, 5, 7: one closed form (j−i)/((i+R)(j+R)) for every gap, built route and closed form agreeing at every depth for all six pairs; the (2,3) pair reproduces the proven strong-EW gap exactly; the NEW convergence g_7 − g_5 = 2/((5+R)(7+R)); every coupling rises strictly toward the One and never reaches it at finite scale — all four forces unify at unison with no tuning, no thresholds, no new fields), and the Grand Lock (every major constant as a closed form over {ONE, b=2, c=3} with the depths counted from c: at c = 3 all land exactly — 34259/250, 1/485, 1/383, 1/95, 27/5, 27/32, 13/12, 127/2, 1/2²⁰; perturb the colour to 4 and every c-dependent constant moves together while the colour-free half-One coupling stays put — the SM's ~26 "independent parameters" are one object seen from many sides).
• The closed inventory, the Smithion chain, and two ancient conjectures made fold-native — the complete particle census (per sector p: p²−1 carriers — the same count that gives 3 weak carriers and 8 gluons forces the NEW 24 penta + 48 hepta; totals counted: 83 gauge carriers, 72 undiscovered, 12 Smithions; the sector count is exactly four and the first prime beyond the ceiling is 11 > 7 — the list ENDS, and any confirmed particle outside it refutes the corpus outright), the Smithion masses (no new scale — every fermion's mass-part is its sector coupling's shortfall from unison, the same rule that gives the electron 1/2 and up quark 1/3 forces penta 1/5, hepta 1/7, ratios to the electron 2/5 and 2/7 — light, confined, sector-only, hence unseen), Collatz as a fold contraction (the odd-even pair multiplies by (3/2)(1/2) = 3/4 — Kleiber's own branching ratio, below the One, so every pair contracts; the cycle 1→4→2→1 walked; every start to 100,000 falls to the 1-cycle, n=27 in its known 111 steps), and prime pairs as fold antipodes (an even number's pair (k, E−k) casts to exactly the One about its half-One — Goldbach verified for all 4,999 evens to 10,000, and the twins counted to the same bound: exactly 205, the published value).
• Turbulence, the violent Earth and sky, free will, and the map of necessity — the Kolmogorov exponents (the cascade branches across the COUNTED three dimensions: the structure exponent is the branching ratio 2/3 — checked equal to the strong coupling, one number two roles — and the spectrum exponent 5/3; a binary cascade verifiably gives the wrong answer), earth & high-energy astro (tipping = a one-step lock crossing, the basins 1/4 and 3/4 folding to the same lock; quakes = the unit power law exact over ten octaves — Gutenberg-Richter b = 1; bursts = atomic release, half-One to the One in one move; ringdown = the fold's own halving, never zero), free will (forward the fold is a function — fully determined, libertarian free will ruled out; backward it is 2-to-1 — distinct pasts, equal presents, one bit lost — so the self cannot pre-read its own next act: determinism plus forced self-opacity, a closure not an open), and the counterfactual map (NECESSARY: 1/2 the unique self-antipode — exhaustive to denominator 300 — base 2 the minimal fold, the depths counted; the master dial moves all NINE c-dependent constants together; CONTINGENT: one bounded label m ∈ {2,3,5,7}; CONVENTION: the unit name — the SM's ~26 dials against the fold's ZERO).
• A solar prediction with nothing local, and the fold as a computing engine — Parker reconnection protons (the cutoff E = 8α² m_p c²: the coupling forced, the eight channels counted, the proton's own scale — 399.714 keV vs Parker Solar Probe's ~400 keV, within 0.1%, with zero local inputs: no density, no field strength, no Alfvén energy), the compact generator (the group law χ_a χ_b = χ_(a XOR b) verified exhaustively, generators DERIVED forward from the rule as fold-projector products — Nim's generator is 8 at 2, 3, and 4 heaps while the field grows 64 → 4096, every state reconstructed exactly: compression is a theorem, not an observation), and the universal exact solver (the chess campaign's retrograde fold, off chess: the subtraction game — 1001 states — and Nim — 512 states — solved and certified against independent closed-form oracles at zero disagreements; the engine was never about chess).
• Number theory native to the fold, the Higgs with no inserted mass, and nature's favourite exponent — fold number theory (five laws, exhaustive: orbit length = ord_q(2) for EVERY reduced p at every odd q to 99; the φ(q)/ord_q(2) orbit census tiles exactly; only the half-One is self-antipodal; even denominators decay their 2-adic valuation onto the odd core; fully-ergodic primes counted to 4000 — 214 of 549, density 0.390 vs Artin's 0.374 — the conjecture renamed: how often is a prime one ergodic mode?), the Higgs sector (the tower's first three rungs 1/2, 1/4, 1/8; Route A off the rungs and Route B through m_H² = 2λv² both close on m_H = v/2 → 123.11 GeV vs measured 125.25, 1.7%; forced λ = 0.125), and the unit power law (only the one-over-rank law halves per rank doubling — square and flat laws verifiably excluded — Zipf's exponent = Gutenberg-Richter's b = the One).
• Phase transitions, the hadron towers, three astronomical orphans, and the end of infinity — critical exponents made exact (β = ν = 1/2, γ = 1, δ = the colour 3; Widom, Rushbrooke, and Fisher all close with ZERO residual, α and η at the floor — the textbook mean-field set, forced), the Regge spectrum (constant tube width = the half-One → M² = M₀² + σJ with spacing checked exactly equal at every rung and at two different anchor pairs — the shape is forced, the tension is the one anchored scale; multiplicity 2^d walked to the ceiling), materials & astro (quasicrystals ARE the prime-5 sector — three standing modes, periodic lattice verifiably forbidden, hence aperiodic; Titius-Bode IS the binary tower — ratios exactly 2; Tully-Fisher's fourth power = d+1, the halo = the relic 27/5), and the finite inventory (towers ≤ 2⁷ = 128 rungs — the floor 1/128 walked; elements ≤ 137; species = 83 carriers + 12 Smithions — the inventory of everything that can exist is a finite count, and every boundary is a standing falsification target).
• The nuclear island, the code's inner structure, biology, and the search strategy — the magic numbers and SMITHIUM (three dimensions + spin-doubling + the colour-3 spin-orbit reproduce all eight known magic numbers 2, 8, 20, 28, 50, 82, 126, 184 with zero fitted coupling; no closure lies between 82 and 126, so the island is forced at SMITHIUM — Sh, Z = 126, N = 184, A = 310 — not the 114/120 of the tuned models), chemistry & biology completion (the 64 codons tile 16 first-two-base boxes with third-base wobble; the 2(2l+1) rule gives s/p/d/f = 2/6/10/14 and the unentered g-block = 18), biology from the fold (aging = transient decay / the Hayflick limit as a 2-adic valuation, the germ line eternal; the neural spike = an atomic threshold fold; cancer = lost descent; ecosystems = bounded-denominator cycles), and applied signatures (the Smithion is electromagnetically dark — no nuclear-recoil channel, so the search is gravitational; the new forces show as confining jets of 24 and 48 carriers plus missing energy; a confining signature beyond prime 7 falsifies the theory).
• The twelve Smithion spectra, and alpha's last two doors closed — the Smithion spectra (the unified coloured invariant I2(c,d) = c/(c(2c^d−1)−1) reproduces the proven quark invariants at c = 3 exactly — cross-checked against the independent sharpened-invariant route, both depths; every spectrum's three sign changes FOUND by grid scan (never chosen) and each root closed by 60 exact halvings; c = 3 validates at 1 : 20.11 : 967 and 1 : 486 : 51140 bare, and the same construction forces all four new spectra — twelve Smithions, three generations each, enormously split towers so only each sector's lightest survives: the relic structure the dark sector requires), and alpha forced, assembly and all (1/α as a pure function of the generators with depths re-counted per mutation — (2,4), (3,3), (2,5) all move the value; the covering volume rebuilt four "plausible" wrong ways — 125, 500, 375, 686 — all miss: the value survives no substitution).
• The mass-weighted LFV table — and the frontier campaign COMPLETE — the LFV spectrum (the generation mass-parts 1/6, 1/2, 5/6 are the checked preimages of the lock; the full physical weights μ→e : τ→μ : τ→e = 1/32 : 5/96 : 5/24 = 3 : 5 : 20 exactly, with the τ channels' 4:1 preference mass-independent — the Belle II fingerprint; the ladder's beta slope p−1 added), and MILESTONE TWO: every standalone frontier engine of the published corpus is now recreated as a clean-room suite or receipted under existing suites (absolute scale, the Millennium positives, the complete periodic table, prime-force phenomenology, the validation harness, and the syntax gate — whose no-zero law the clean-room enforces at the stronger VALUE level). Proof corpus + frontier engines = the entire published derivation surface of SFTOM, in exact arithmetic, forced from the One.
• The discovery engine remade, and the chess theorems off chess — the USDE under the law (the autonomous sweep, exact and deterministic: the standing-mode census counted two independent ways — 5,022 members — closure under all four sector forces PROVEN member by member, the orbit partition tiled, the emergence family swept across the sectors; at m = 3 the swept invariants ARE the forced lepton 1/485 and Koide 2/3 as exact identities, at m = 5, 7 the family forms the predicted 1/6249 and 1/33613; the couplings, the mixing partition, the dark shares, the full cosmic budget, the reactor angle, and the floor all fall out of the sweep — no floats, no tolerance dial, no seed, no null model, no statistical annotation: a deterministic engine has no chance to correct for, and it runs in 0.02 seconds), and the certified game theorems (the chess campaign's twin-pair law and vanishing law re-proven exhaustively on independent solved fields — every Nim coefficient equals its swapped twin, every odd-overlap subtraction coefficient is exactly zero, half the spectrum provably empty: the storage corollary by count — the pipeline, not the board, is the instrument).
• Alpha exact, and the fold plays chess — the terminal fine structure (the self-similar ladder is counted and it TERMINATES — one cube direction promoted per order, 5³ → 5²·7 → 5·7² → 7³, four rungs = colour + 1, then no successor exists: 1/α = 503846395469/3676744786 = 137.035999177180855… exactly, 0.009σ from CODATA; the digits at 2×10⁻¹¹ called in advance — …177181 — and the live 5.4σ Rb/Cs discrepancy called to resolve at 137.0359991772; see PREDICTIONS.md), and the fold chess bot (complete legal chess — castling, en passant, all four promotions, mate — with zero fitted parameters: rules certified at FOUR published perft positions — the start 8,902, Kiwipete 97,862, the en-passant-pin endgame 43,238, four-way promotions 9,483 — zero disagreements at every depth; material counted from the board's own geometry; the starting position EXACTLY the half-One; minimax through the fold's antipode; and in refereed whole games with python-chess independently validating every move it emits: 10–0 vs a random legal mover and 5 draws + 1 loss vs Stockfish at its minimum exposed strength, zero illegal moves (tools/MATCHES.md) — the fold does not just derive; it plays).

The full, ordered, run-it-yourself derivation of all of it — every step, all 409 test suites (2,693 forced checks, including the trace-to-the-One and the CODATA comparison) — is in OneFoldMaster.md, and provable in one run with make -C verify prove.

---

The foundation is a self-proven theorem

The model has zero free parameters and zero axioms. Its single foundation is the machine-checked, self-proven theorem there is no nothing. The foundation proves itself, and the proof is machine-checked in two parts.

The fold itself is forced (constants/forced_fold_theorem.ep). Build maps from only x and the One with +, −, ·, cast_out and no other literal: every constant you can build is a positive integer, and cast_out sends every one to the One — so no continuous free parameter is even expressible, and the candidate operations form a discrete, listable space. The engine begins from the four primitive self-maps {identity, square, constant, fold}, then mechanically generates every ordered composition through size three: 4 + 16 + 64 = 84 forms. It runs every form and finds the Fold is the unique least-size generator; the others are static, collapse, contract, or are nonminimal compositions. forced_unique halts the engine if any equal-size rival qualifies. The exact statement is conditional on this declared generated composition grammar, with larger sizes retained as an extension route.

The One then follows (constants/the_axiom_is_a_theorem.ep):

1. There is no nothing. Zero is forbidden; a value must lie in (0, 1].
2. The displaced ground. The ground cannot be zero — it is the unique value whose double is the One (x + x = One), solved to the half-One 1/2.
3. The One, forced. The ground folds up to the One (fold(1/2) = 1), and the One is the fold's own fixed point (fold(1) = 1) — the unique unison.

So "the One and its fold" is not posited. It is the only consistent foundation, and every link in that chain is machine-checked.

---

Verify it yourself

The proofs are written in ErnosPlain, an exact-arithmetic language; each module is plain, fully spelled-out English with a docstring on every function. You can verify everything two ways (see verify/):

With nothing but a C compiler — no ErnosPlain needed. The verify/ directory holds the ErnosPlain compiler's own output as self-contained C (standard headers only), so any cc/gcc/clang builds it. The one-line way to prove the whole thing:

make -C verify prove        # THE one command: proves absolutely everything
make -C verify check        # or: build & run every proof, one ok / FAIL line per suite
make -C verify online       # cross-check forced values against LIVE-fetched NIST CODATA
make -C verify uniqueness   # the 1/α assembly is the ONLY 9-digit hit over the smooth space

make prove runs every proof in a single pass and prints, for each constant and scale: its forced value, the trace back to the One (test_trace_to_the_one: self-proven foundation theorem → One → counted generators → depths → the constant), and the external measurement it matches (test_codata_comparison: every forced value vs CODATA / PDG / Planck, through a sealed boundary a measurement can never cross into a derivation). It ends with EVERYTHING FORCED, DERIVED, COUNTED, AND VERIFIED — traced to the One, and checked against external CODATA / PDG / Planck measurement (currently 409 suites, 2,693 forced checks, 0 failures).

make online closes the "you typed the measured numbers in yourself" objection: it derives the forced values from the two generators (2, 3) and fetches the measured ones live from NIST's CODATA table — 1/α, the proton/electron mass ratio, and the electron g all land on the downloaded values, nothing measured stored locally.

Reproduce the headline results (theory + the UnisonAI engine): four commands. fold_ai/REPRODUCE.md — the theory prover (409 suites / 2,693 forced checks), the LLM-presence suite on public GPT-2 (13/13 tensors, 39/39 registered checks, weights auto-fetched on a fresh clone), the pre-registered loud-subspace transfer test (verdict SUPPORTED), and the engine's end-to-end verification (36/36), each with its expected output and runtime.

From the ErnosPlain source — the compiler is bundled. The ErnosPlain compiler ernos (a small, dependency-free Rust crate, MIT-licensed) is included in compiler/ so you can compile and edit the .ep sources yourself. Build it once, then either rebuild the C (verify/build_from_source.sh) or run a proof directly:

cargo build --release --manifest-path compiler/Cargo.toml   # builds compiler/target/release/ernos
compiler/target/release/ernos tests/test_fine_structure_constant.ep
./tests/test_fine_structure_constant

(The compiler also lives in its own repo, MettaMazza/ernos-programming-language; the copy here is bundled so the theory and the language that expresses it travel together.)

Each test prints the forced value and checks it exactly. Measured numbers appear only on the comparison side, inside the tests — never in a derivation, and that is enforced, not merely observed: a measured value is the distinct Measured type (foundation/measured_values.ep), which the Fraction-only derivation primitives cannot consume, so feeding a target into a forcing is a compile error — the engine will not build (with a runtime backstop, forbid_target_input, that halts too). The enforcement layer (foundation/enforcement.ep) also halts the engine on any value that is fitted (forced_to_be), chosen among equals (forced_unique), or outside the domain (require_in_domain). And the assembled forms are guarded too: for each constant the candidate space is generated by a stated grammar — every assembly of the forced ingredients with + − · / up to a size bound (foundation/assembly_enumeration.ep) — and the engine proves the forced value has no simpler assembly. A single fitted number, or a needlessly complex form, voids the engine.

---

FoldBot Chess computational-proof project

FoldBot is a complete legal chess engine whose material quantities are counted from current board geometry and whose evaluations are exact rational shares of the One. Its perft and restoration anchors remain exact, and the pinned ladder secures Stockfish 1700 at 54.2% and 1900 at 62.5%. The v20/current calculation preserves every legal root move to common depth, agrees with the sequential engine on 4/4 moves and 4/4 exact values, derives worker messages from the supplied ceiling, and addresses tables from their live capacities. Current and pinned v20 agree on all 36/36 real 2100 positions at ceiling 3; current source completes the same panel at depths 8–11. The next state is a lossless calculation improvement measured on that exact panel, followed by the Maria-authorized 2100 run. The secured 1900 victory and current calculation evidence establish a constructive path and no theorem-derived wall.

Fold Go computational-proof project

Fold Go secures exact legal-position censuses 1, 57, 12,675, 24,318,165 through 4×4, exact empty-board values through 2×2, and two independently replayed 2–0 GNU Go 9×9 batches. The competitive state now binds board, mover, complete positional-superko history, and previous-pass status; typed transposition search agrees with the no-cache all-actions reference on 134/134 sealed states including 24 pass-pass terminals. Empty-board search retains every legal dihedral orbit and quiescence retains every legal active front. Matched 5×5 receipts show exact depth-sensitive selections B5 → B4 and pass → A3. The next state is the Maria-authorized 9×9/full-board campaign and continued development through the strictly-over-50-percent victory criterion. The exact and applied surfaces establish an executable route and no theorem-derived wall.

Fold Protein computational-proof project

Protein Material Architecture V1 achieves Blind Predictive Super Parity across 24 sealed whole-structure tests. Four preregistered panels record 0.9255486262 median TM_repo / 0.7833590149 Å median Cα RMSD95 / 0.9882113352 best whole-structure TM_repo; 15/24 structures equal or outperform AlphaFold's reported 0.96 Å CASP14 median Cα RMSD95. Every complete state path and PDB is sealed before its target coordinates are opened. The original complete 76-residue material execution measures 0.9891211351 TM / 0.2608575408 Å Cα dRMSD / 0.3261459535 Å Kabsch Cα RMSD. It restores all 576 exact states at every residue, applies the SFT three-residue window, two-residue overlap and One-residue advance to sequence/generated-material frames, executes 43,776 raw state trials, and uses zero candidate orderings, weights or fitted parameters.

The supporting engine-checked 3D law forces the canonical right-handed alpha-helix angles (−60°, −45°) from exact circle coordinates (−1/6, −1/8) and the beta-sheet angles (−120°, +135°) from (−1/3, +3/8). The sequence programme completed target-isolated, pre-comparison-sealed executions through all 76 ubiquitin residues and advanced complete-domain assembly through V45. Its preserved applied evidence includes local PSD at 0.9997464589 TM / 0.0077621385 Å dRMSD, a V43 whole-chain frontier at 0.1797422881 TM, and a V44 connected One-cycle distance frontier at 5.8944799638 Å dRMSD.

This is empirical blind prediction because an explicit observationally derived law produces sealed, falsifiable structures before independent experimental measurement. Predictive parity is one component of Super Parity. The combined result adds first-principles derivation, machine-checked provenance to the One, zero trained and fitted parameters, transparent state-by-state execution and public reproduction. One researcher produced the programme on one Mac Studio in approximately one week divided among four SFT computational-proof programmes. AlphaFold is the industrial black-box accuracy comparator; its reliability does not supply the inspectable law or proof delivered here. Execution across unrelated fold families can extend demonstrated reach without postponing or redefining this achieved result. See the authoritative standalone paper From_One_Theorem_to_Blind_Protein_Structure_A_Zero_Parameter_Computational_Proof.md, concept DOI 10.5281/zenodo.21482127, and the canonical Fold Protein repository. The two earlier Protein papers are retained as superseded development artifacts.

---

UnisonAI computational-proof project

UnisonAI is the active artificial-intelligence computational proof of this corpus. Its canonical version-controlled development repository is MettaMazza/UnisonAI; reviewed established releases are synchronized back here.

The native architecture uses zero trained parameters. Its promoted mechanisms are tracked as directly forced, forward-forced, or constitutionally re-derived; disclosed engineering quantities record open development status and are not theoretical walls. The current standalone structural verifier passes 21/21 checks, while the main SFT corpus supplies the closed coherence, contextual-integration, and generation-selection laws in Steps 322–324.

The current native v4 causal store contains 649,917 role-bound pairs, 11,140,970 assistant targets, 22,415,744 counted Q/K relations, 96,721 value vectors, and 81,111,826 deep FFN addresses. Exact packed serving preserves every row and byte-identical output while reducing the fixed probe from approximately 59 GB to 559,185,920 bytes and from 1,514.370707 to 11.407350 seconds. Prompt-admitted induction produces 4/4 matched continuity and 8/8 transfer while removing full-statement copies.

The complete v5 position relation is now sealed: 277,583,049 observations, 212,395,127 unique canonical entries, 4,369,646,672 canonical bytes, and 5,097,483,048 packed bytes. Runtime integration of position-owned values and semantic FFNs, free multi-turn composition, reward-conditioned learning, and benchmark victory are the next state. Counted prediction has already beaten matched trained twins at character scale (1.2891 vs 1.8878) and word scale (3.1907 vs 3.4292). The complete transformer organ map and v5 relation establish a constructive continuation and no theoretical wall on full conversational generalisation. Supporting local models are scaffolding and judges, not native Unison parameters. See papers/UnisonAI_The_Forced_Language_Architecture.md and the standalone repository for the synchronized release boundary.

---

foundation/   the One, the fold, exact integers and fractions, the counted
              generators, the enforcement that forbids fitting or choosing, and the
              assembly enumerator that generates each candidate space from a rule
constants/    one module per forced result (the table above and more)
tests/        one test per module — run them; read the `ok` lines
verify/       committed C + Makefile: reproduce everything with just a C compiler
OneFoldMaster.md   the complete, ordered, auditable derivation, step by step
STANDARDS.md       the two rules every module is held to
AGENT.md           for an AI or skeptic: why there is no degree of freedom to choose

---

Smithian Fold Theory — Maria Smith, Ernos Labs. The constants of nature are not free. They are forced.