# The Metastable Solitude — validated component pilot

The complete inherited numerical workflow reproduces, and independent controls reinforce its main limitations. This release does not demonstrate a matched, observationally allowed physical history or an explanation of missing peers. It supplies an executable uv project, frozen inputs, run records, a revised essay and seven public interactive figures (nine static gallery exports).

## What was run

E00 verifies all 49 legacy hashes and all 20 entries in the handoff manifest, reruns the eight original tests, and regenerates the entire legacy workflow in a separate working copy. All 16 CSV tables match at the declared per-table tolerances. This includes 135 environmental network grid cases, 15 single-field bounces, three two-field bounces and the radial fluctuation check. The regenerated JSON diagnostics are saved, but the automatic historical array comparison covers CSVs; JSON identity is not claimed.

All 98 REACLIB coefficients (14 seven-coefficient fit components) were independently parsed from the pinned upstream C++ files; both Git blob hashes agree with the archive. The historical FY05/NAC2/CO10/IL10/ST08 rates are not a newly evaluated reaction library. The source's exponent floor is irrelevant to the reported helium-burning pilot; it is not applied to manufacture large-temperature-range equivalence.

The supplied directory differs from the handoff's stated structural success: an extra root `REPORT.md` links to absent `figures/` files and is not covered by its manifest. The original is preserved. `handoff-validator.log` records the failure, while the legacy ZIP and its independent numerical reproduction pass. The AME2020 mass endpoint returned HTTP 403; no substituted mass values or evaluated-threshold plot was fabricated.

## Results and counterexamples

At 200 million K and 10,000 g/cm³, the independently integrated six-species parcel leaves carbon mass fractions 0.919536, 0.760293 and 0.382765 for imposed shifts −39.9, 0 and +39.9 keV after 99.9% helium consumption. Corresponding elapsed times are approximately 15,096, 25,287 and 57,605 years. These are restricted constant-condition parcels, not stellar lifetimes or ejected yields.

Let the total triple-alpha coefficient ratio at fixed temperature be r. Changing density to ρ/r multiplies every forward flux by 1/r, so the same composition path is traversed in r times the original duration. For +39.9 keV, r≈0.0987562 and the adjusted density is about 101,259 g/cm³. The independent time solver and the inherited depletion-coordinate solver agree. Equal-time outputs are never extrapolated beyond the saved stopping event. A negative-shift trajectory stopped sooner is explicitly marked unavailable at the later baseline time.

The E14 finite-difference experiment projects the five final heavy-species mass fractions against the density direction. Its relative residual decreases to about 2.7×10⁻¹¹ as the derivative step is refined. No measurement covariance or significance is inferred. Elapsed time could break this composition-only degeneracy if age were independently constrained. Screening, reverse reactions and actual stellar adjustment can change it.

The selected scalar–Higgs potential has stable classical minima and a phase-energy difference of 1.25×10⁷ GeV⁴. Its O(4) bounce action reproduces at about 399.823. The radial l=0 sector contains one negative mode among the checked low eigenvalues; the translation mode approaches zero on refinement. Neither these checks nor an assumed prefactor provide a cosmological lifetime. The energy difference is an obstacle to interpreting the benchmark as otherwise ordinary regions; it is not a universal dark-energy-ratio exclusion.

E05 declares two matching scenarios. A fixed-QCD tree-level diagnostic holds Λ fixed. A one-loop common heavy-threshold scenario gives Λ/Λ_A=(h/h_A)^(2/9), m_e/Λ and light-quark/Λ ratios scaling as (h/h_A)^(7/9), and G_FΛ² as (h/h_A)^(-14/9). Fixed high-scale strong coupling and Yukawas are assumptions. QED threshold matching and higher orders are absent; alpha is explicitly held fixed. Neither map silently supplies the independent alpha changes in E01. An even sign-symmetric potential with even matter couplings gives identical local properties in ±s phases.

The M2 control solves ∇²s=m_in²s inside a uniform sphere and ∇²s=0 outside, with continuous value and derivative and s(∞)=s_inf. Analytically s(0)/s_inf=sech(m_in R); an independent boundary-value solver agrees. At m_in R=10 the core fraction is about 9.08×10⁻⁵. This linear limiting equation illustrates how screening can remove the effect in the material of interest. It is not a nonlinear symmetron, an actual stellar profile, a force constraint or a second metastable chemical phase.

Hydrogen eigensolves at three mesh spacings show second-order convergence and agreement in rescaled Hartrees when alpha or electron mass is changed. The finest-grid ground-state absolute error is below 4×10⁻⁶ Hartree. Nuclear recoil, relativity, molecules and adaptation are absent. The visual shows computed atomic levels; it does not substitute a hand-drawn molecular potential.

For homogeneous Poisson fronts in flat space, p₀=exp[−πΓv³t⁴/3] and u(z)=z−z³/4+z⁴/16 for z≤2, with u=1 beyond that. Old-phase conditioning gives p₀^u. New-phase inclusion–exclusion is evaluated using expm1 to remain stable even at a marginal fraction of 10⁻²⁰. Quadrature and 250,000 seeded samples independently check the 3D-space-plus-time cone overlap. The Monte Carlo validates overlap volume, not direct sampling of extremely rare surviving observers.

At marginal fraction 10⁻⁶ and z=10⁻⁵, the conditional probabilities are 0.9998618544 (old) and 0.9999900000 (new). These are shared-phase probabilities, not life probabilities. The luminal point-origin causal test follows the triangle inequality. In a collinear N=0, S=1, O=3 control, a photon emitted as the front reaches S arrives with a light-speed direct front at t=3; at v=c/2 it arrives at t=4 before the front at t=6. The slower speed is stipulated, not dynamically predicted.

## Protocol coverage and exact next work

`validated_within_scope` below applies to the described component only, not to completion of every stage in the original protocol. Planned work is distinguished from missing physical inputs and missing resources.

| Task | Lifecycle | Completed scope | Next calculation or input |
|---|---|---|---|
| E00 | validated_within_scope | Full inherited numerical workflow, 49 ZIP checksums, eight tests and 16 CSV comparisons. Historical coefficients independently parsed upstream. Extra root REPORT.md breaks the handoff's structural validator. | Reproduce a second evaluated rate library and document normalization differences. |
| E01 | validated_within_scope | Fixed-temperature exponential, zero-shift and derivative controls; historical total includes nonresonant terms. Local alpha derivative is a literature input. | Acquire consistent evaluated masses and levels, resolve electronic binding corrections and correlated nuclear response. AME endpoint returned HTTP 403. |
| E02 | planned | Only the inherited partial E1 denominator is reproduced. Full E1/E2 implementation has NOT been completed. | Implement S03 Appendix A interference, Coulomb/Whittaker integrals and regulators; validate E1/E2 baselines, then thermal quadrature. Matched effective-interaction derivatives require external nuclear calculations. |
| E03 | validated_within_scope | Independent time-domain six-species integration, baryon conservation, equal-time/depletion comparisons and exact density compensation. | Add reverse reactions, screening and channels with baseline reduction checks, then use correlated rates. No time-varying particle masses are inserted here. |
| E04 | blocked_resources | MESA and its SDK/compiler are not installed; no stellar track, yield or custom-rate adapter is claimed. | Install a pinned MESA/SDK toolchain (official installation: about 2 GB download, 20 GB installed; https://docs.mesastar.org/en/26.4.1/installation.html). Reproduce one 15-solar-mass baseline with documented inlists before any shift. |
| E05 | validated_within_scope | Classical Hessians, distinct Higgs backgrounds, equal-background null, and two declared dimensionless matching scenarios. No single matched viable model. | Match heavy thresholds beyond leading order and QED running, nuclear responses, renormalized couplings and EFT cutoff; retain all predictions under one model ID. |
| E06 | validated_within_scope | Inherited flat-space classical two-field bounces, convergence and radial fluctuation spectra rerun. Prefactor, radiative, thermal, gravitational and population calculations remain open. | Independent multi-field solver and renormalized effective potential, then justified gravitational and population histories. Do not use a dark-energy ratio as an exclusion criterion. |
| E07 | validated_within_scope | Linear uniform-sphere screening control independently checked by a BVP solver. This is a limiting equation, not a nonlinear symmetron or stellar interior. | Choose a complete M2 potential/coupling and solve nonlinear radial profiles with actual stellar densities and matching force predictions. |
| E08 | validated_within_scope | Exact flat-space conditioning, independent cone quadrature/Monte Carlo, stable rare-new-phase limits and luminal/slower causal controls. | Generate dynamical 3D histories, observer-history selected realizations and past-light-cone observations with model-derived speeds; no emergence or detection likelihood exists. |
| E09 | validated_within_scope | Independent nonrelativistic infinite-nuclear-mass hydrogen eigensolves, second-order mesh convergence and rescaled collapse. | Compute H2 then the molecular probe set, basis convergence and non-scaling corrections. No chemistry or biological function is simulated; transition work must be included in dynamic extensions. |
| E10 | planned | Published absorber summaries remain literature context; no synthetic spectrum ensemble or raw-spectrum fit. | Freeze transitions and primary sensitivity coefficients, inject realistic blends/calibration errors, test recovery/coverage, then fit the released absorber. |
| E11 | blocked_missing_input | M3 fossil enrichment remains a hypothesis. No stellar ejecta and delay kernels are available from E04. | Obtain matched total/net yields and release times; validate gas mass balance, delay kernels, mixing and ordinary-history degeneracy before catalog fitting. |
| E12 | planned | No integrated clock, force, collider or BBN likelihood was constructed. Remote minima are not local derivative bounds. | Map the same specified theory point into observables with range, composition, screening and epoch; calibrate synthetic recoveries before inference. |
| E13 | planned | Optional M4 branch is not implemented. | Select a definite canonical/noncanonical model and validate localized motion, energy losses, boundary effects and dimensional lifetime. |
| E14 | validated_within_scope | Finite-difference signal projected against density variation at fixed depletion. Composition-only residual tends toward zero; this has no significance calibration. | Complete a compatible field/nuclear/stellar/observation benchmark. The pilot establishes no unified allowed point and no Fermi explanation. |

## Frozen evidence map

| Task | Run |
|---|---|
| E00 | `20260927T015443056558Z-E00` |
| E01 | `20260927T015330656008Z-E01` |
| E03 | `20260927T023342111171Z-E03` |
| E05 | `20260927T015335289373Z-E05` |
| E06 | `20260927T015443056558Z-E00` |
| E07 | `20260927T015335313304Z-E07` |
| E08 | `20260927T015443415499Z-E08` |
| E09 | `20260927T015444052071Z-E09` |
| E14 | `20260927T015444724011Z-E14` |
| E00 upstream transcription | `20260927T022107855063Z-E00-sources` |

| Claim | Current conclusion | Component runs |
|---|---|---|
| C01 | Two classical stable backgrounds and the inherited bounce reproduce. Corrected theory and observations remain untested. | 20260927T015443056558Z-E00, 20260927T015335289373Z-E05 |
| C02 | Fixed-temperature rate response checked; full capture and matched nuclear thresholds remain incomplete. | 20260927T015330656008Z-E01, 20260927T015335289373Z-E05 |
| C03 | Independent parcel integration confirms changes and exact density compensation; no distinguishable stellar yield is established. | 20260927T015330656008Z-E01, 20260927T023342111171Z-E03, 20260927T015444724011Z-E14 |
| C04 | Use only the demonstrated functional class; do not describe a region as universally lifeless. | No new supporting result |
| C05 | The reproduced bounce and radial modes are not a computed cosmic lifetime or phase population. | 20260927T015443056558Z-E00, 20260927T015443415499Z-E08 |
| C06 | The chosen model's phase-energy difference is large; gravitational evolution is not solved and no general exclusion follows. | 20260927T015335289373Z-E05, 20260927T015443056558Z-E00, 20260927T015335313304Z-E07, 20260927T015444724011Z-E14 |
| C07 | Measurements constrain specified systems and interactions; no unified allowed point has been demonstrated here. | No new supporting result |
| C08 | Global rarity does not imply local solitude in the tested Poisson-front history; causal visibility further restricts the luminal case. | 20260927T015443415499Z-E08, 20260927T015444724011Z-E14 |
| C09 | A limiting spherical screening equation suppresses the core field; nonlinear matched stellar and force calculations remain absent. | 20260927T015335289373Z-E05, 20260927T015443056558Z-E00, 20260927T015335313304Z-E07 |
| C10 | Fossil enrichment is a proposed route to historical evidence. | No new supporting result |
| C11 | Transition damage is a question to calculate, not an assumed outcome. | No new supporting result |
| C12 | The Fermi connection is an additional unresolved claim. | No new supporting result |
| C13 | Hydrogen spectra collapse in rescaled units, with mesh convergence; molecular and non-scaling effects remain uncomputed. | 20260927T015444052071Z-E09 |
| C14 | The parcel's composition response lies in its density nuisance direction. No observed signal or full observational inference exists. | 20260927T015444724011Z-E14 |
| C15 | Oscillatory persistence is an optional different hypothesis class. | No new supporting result |

## Reproduction and numerical limits

From `research/metastable`, install the frozen project with `uv sync --frozen`, then use `uv run --offline --frozen python run.py smoke`. The full restart sequence is `audit --full`, `check-sources`, `run pilot`, `validate`, `render`, `report`, each following `uv run --offline --frozen python run.py`. Explicit network retrieval is isolated in `acquire`; no core test downloads anything. Repeated runs get new IDs; failed records remain.

Run manifests contain commands, Python/platform/package versions, configuration and input hashes, code hashes, seeds, checks and elapsed time. The legacy host used Python 3.13.5; this pilot uses 3.12.12 on Apple Silicon with the same pinned numerical package versions. CPU time in a parent manifest excludes subprocess CPU; legacy per-stage wall-time output is retained. No background execution is promised.

Matplotlib exports SVG/PNG, source tables and JSON captions/provenance. The essay's parcel animation uses convex interpolation of saved samples (sample times span a logarithmic grid to resolve early burning; playback is explicitly linear time or depletion). Its mass fractions remain nonnegative and baryon-normalized. The causal animation evaluates its declared analytic geometry. Both start paused and stop at the end. Euclidean bounces are static inspections, never time-lapse crossings. Every public figure has a static export and a textual account. The gallery includes diagnostics omitted from the essay.

No parameter scan here is a prior over universes. No universal biological threshold, full nuclear capture, fully matched star, molecular function, renormalized gravitational history, observational fit or detected anomaly is claimed.