Earn the chronometric evidence frontier: adversarial repair v2

.github/workflows/chronometric-frontier-evidence-v2.yml

@@ -0,0 +1,68 @@
+name: Chronometric frontier evidence v2
+
+on:
+  push:
+    branches:
+      - chronometric-frontier-evidence-v2
+    paths:
+      - .github/workflows/chronometric-frontier-evidence-v2.yml
+      - chronometric-emergence/frontier-evidence-v2/src/**
+      - chronometric-emergence/frontier-evidence-v2/tests/**
+      - chronometric-emergence/frontier-evidence-v2/pyproject.toml
+      - chronometric-emergence/frontier-evidence-v2/Makefile
+
+permissions:
+  contents: write
+
+jobs:
+  evidence-repair:
+    runs-on: ubuntu-latest
+    timeout-minutes: 120
+    steps:
+      - name: Check out repair branch
+        uses: actions/checkout@v4
+        with:
+          ref: chronometric-frontier-evidence-v2
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+          cache: pip
+          cache-dependency-path: chronometric-emergence/frontier-evidence-v2/pyproject.toml
+
+      - name: Install evidence package
+        working-directory: chronometric-emergence/frontier-evidence-v2
+        run: python -m pip install -e '.[test]'
+
+      - name: Run full independent evidence suite
+        working-directory: chronometric-emergence/frontier-evidence-v2
+        env:
+          CHRONOMETRIC_REPO_ROOT: ${{ github.workspace }}
+        run: python -m frontier_evidence.run_all --repo-root "${GITHUB_WORKSPACE}"
+
+      - name: Run tests from a clean checkout layout
+        working-directory: chronometric-emergence/frontier-evidence-v2
+        env:
+          CHRONOMETRIC_REPO_ROOT: ${{ github.workspace }}
+        run: pytest -q
+
+      - name: Commit generated evidence
+        run: |
+          git config user.name "Angus Muffatti"
+          git config user.email "130276587+terrorproforma@users.noreply.github.com"
+          git add chronometric-emergence/frontier-evidence-v2/results \
+                  chronometric-emergence/frontier-evidence-v2/docs/FRONTIER_EVIDENCE_REPORT.md
+          if ! git diff --cached --quiet; then
+            git commit -m "Generate adversarial frontier evidence v2 [frontier-generated]"
+            git push origin HEAD:chronometric-frontier-evidence-v2
+          fi
+
+      - name: Upload evidence bundle
+        uses: actions/upload-artifact@v4
+        with:
+          name: chronometric-frontier-evidence-v2
+          path: |
+            chronometric-emergence/frontier-evidence-v2/results
+            chronometric-emergence/frontier-evidence-v2/docs/FRONTIER_EVIDENCE_REPORT.md

chronometric-emergence/frontier-evidence-v2/AUDIT_RESPONSE.md

@@ -0,0 +1,22 @@
+# Response to the independent audit
+
+The audit's central criticism is accepted: the project often treated identity checks and regression fixtures as if they were independent physical evidence. The repair does not contest that by rhetoric. It changes the evidence architecture and reruns the load-bearing layers.
+
+## Findings accepted without qualification
+
+- The v1.4 collinear mass combination was structurally wrong.
+- The v1.3 finite “RG completion” was chosen to cancel the scale logarithm and was not derived from an anomalous-dimension calculation.
+- The v1.9 package was a specification/preflight bundle, not a 3PI evolution solver; “pilot authorized” was false.
+- The v1.7/v1.8 longitudinal Ward checks were largely true by construction.
+- Stored v1.5 values were regression fixtures unless explicit recomputation flags were used.
+- The v0.7 `f_a >= M_P` compactness corollary was applied outside its domain on the low-`f_a` ridge.
+- The prompt daughter in the v1.2 cascade had no demonstrated thermalisation channel.
+- Sandbox-specific paths prevented clean-checkout reproducibility.
+
+## Audit numbers treated as hypotheses until reproduced
+
+The audit reported several precise downstream numbers. This package independently recalculates the transport shifts, prompt-daughter dark radiation, compactness, off-shell model spread, and UV-quality requirement. The audit's values are not copied into acceptance tests.
+
+## Scientific consequence
+
+The narrow formal core and leading one-loop threshold interpretation survive. The cosmological and non-Abelian-HPC superstructure does not retain frontier status merely because a preflight script passes. Negative results and blockers are now first-class outputs.

chronometric-emergence/frontier-evidence-v2/AUTHORS.md

@@ -0,0 +1,5 @@
+# Authors
+
+**Angus Muffatti** — research programme, models, calculations, manuscripts, and this evidence-repair package.
+
+AI assistance was used for derivation, implementation, testing, and editorial work. Every result in this folder is classified by evidence type; no AI-generated statement is treated as independently verified merely because code reproduces it.

chronometric-emergence/frontier-evidence-v2/ERRATA.md

@@ -0,0 +1,13 @@
+# Errata and claim corrections
+
+1. **v1.3 RG completion — retracted.** The displayed finite term was chosen to cancel the explicit matching-scale logarithm. The cancellation is an identity pin, not a derived anomalous-dimension result. The natural-scale hard matching value may be retained; the claimed RG closure may not.
+2. **v1.4 direct AMY numbers — deprecated.** The collinear mass combination was implemented with the wrong structure. This package recomputes the diagnostic with the corrected combination and treats v1.5 as the principal LPM anchor.
+3. **v1.5 default verifier — relabelled.** Stored high-resolution and scan values are regressions unless `--recompute-high --recompute-scan` is used.
+4. **v1.7–v1.8 Ward/STI tests — relabelled.** Line-integral longitudinal vertices are construction pins. They are not evidence that a finite non-Abelian truncation preserves the full dynamical STI/Nielsen system.
+5. **v1.7–v1.8 off-shell kernels — model family only.** Their on-shell anchors are useful; their arbitrary off-shell shapes are ansätze. The inter-version spread is reported as model discrepancy, not a calculated uncertainty bar.
+6. **v1.9 pilot authorization — withdrawn.** The package contains specifications, ledgers, preflight checks, and loaders but no 3PI/Kadanoff–Baym evolution engine.
+7. **v0.7 screening theorem — domain restricted.** Its `f_a >= M_P` corollary cannot be applied to the low-`f_a` cosmological ridge. Compact objects are recalculated here.
+8. **v1.2 prompt daughter — original assumption unsupported.** A sterile/free-streaming daughter overproduces dark radiation. This package supplies an exact-replica, gauge-charged daughter repair and a rate calculation.
+9. **v0.2 speed/gap relations — model-specific.** They are tree-level consequences of the minimal one-mediator EFT, not universal laws.
+10. **v0.4 high-scale scalar benchmark — withdrawn as a UV-stable benchmark.** The one-loop quartic running found later requires threshold stabilisation or a two-loop vacuum analysis.
+11. **Precision.** Reported digits now follow numerical resolution and theory uncertainty; symbolic identities may retain exact forms.

chronometric-emergence/frontier-evidence-v2/EVIDENCE_POLICY.md

@@ -0,0 +1,16 @@
+# Evidence policy
+
+The former project matrices compressed unlike checks into the word `PASS`. This package forbids that practice.
+
+| Class | Meaning | What it does **not** establish |
+|---|---|---|
+| `IDENTITY_PIN` | An algebraic identity or construction is encoded consistently. | Physical correctness of the construction. |
+| `REGRESSION` | Current code reproduces an earlier stored output. | Independent recomputation. |
+| `INDEPENDENT_RECOMPUTATION` | A separate implementation reaches the result from declared inputs. | Continuum convergence or empirical truth. |
+| `CONVERGENCE` | The observable stabilises under resolution, timestep, cutoff, or memory changes. | Correctness of the underlying model. |
+| `EXTERNAL_BENCHMARK` | Agreement with an analytic or published result not generated by the tested code. | Validity outside the benchmark domain. |
+| `PREDICTION` | A genuinely fallible calculation produces a model consequence not built into the test. | Experimental confirmation. |
+| `BLOCKED` | A load-bearing issue is unresolved. | Permission to silently assume the issue away. |
+| `RETRACTED` | A prior claim is withdrawn. | That downstream numbers remain valid. |
+
+A frontier claim requires at least one independent recomputation, one convergence study, and one external benchmark for each load-bearing numerical layer. Identity pins and regressions remain useful software tests, but they may not be cited as physical validation.

chronometric-emergence/frontier-evidence-v2/FRONTIER_GATES.md

@@ -0,0 +1,32 @@
+# Frontier gates
+
+## Gate A — formal core
+
+The universal-clock factorisation theorem and the definition of chronometric shear must be stated with their exact hypotheses. The rank diagnostic may infer the number of fields only under full-rank sensitivity and independent-excitation assumptions.
+
+## Gate B — QCD transmission
+
+`2/27` is retained strictly as the leading one-loop threshold coefficient. Higher-order matching and hadronic sensitivities must carry an uncertainty band.
+
+## Gate C — transport
+
+The v1.4 numerical table is deprecated. A corrected mass combination, convergence run, Bethe–Heitler limit, and external normalisation benchmark are required. The v1.5 implementation is the current anchor, not a proof of the complete leading-order portal rate.
+
+## Gate D — real-time evolution
+
+A real two-time calculation must evolve from equations rather than reload stored answers. Timestep and memory-window convergence, commutator normalisation, FDT/KMS, and energy diagnostics are mandatory. The scalar open-system unit tier in this package satisfies this software gate. It does not substitute for the non-Abelian 3PI solver.
+
+## Gate E — cosmology and compact objects
+
+The prompt daughter in the fermionic cascade must thermalise through an explicit interaction. The low-`f_a` ridge must be evaluated for compact objects. Neutron-star conversion and binary constraints remain model blockers unless a demonstrated screening or alternate cosmological-selection mechanism is supplied.
+
+## Gate F — UV quality
+
+The elementary global-field implementation fails the required quality bound. A local gauge/deconstructed completion must suppress the first allowed winding operator below the declared vacuum potential.
+
+## Current authorization
+
+- Scalar two-time unit tier: **authorized and implemented**.
+- Corrected transport evidence run: **authorized; results generated by CI**.
+- Full PT/BFM three-loop 3PI pilot: **NOT AUTHORIZED** until an actual evolution engine exists and Gates C–F are resolved at the level relevant to the claimed observable.
+- Publication of the narrow formal/threshold note: **conditional** on specialist novelty review and independent source audit.

chronometric-emergence/frontier-evidence-v2/Makefile

@@ -0,0 +1,12 @@
+PYTHON ?= python3
+REPO_ROOT ?= ../..
+
+.PHONY: reproduce test clean
+reproduce:
+	$(PYTHON) -m frontier_evidence.run_all --repo-root $(REPO_ROOT)
+
+test:
+	pytest -q
+
+clean:
+	rm -f results/*.json results/*.csv results/*.npz docs/FRONTIER_EVIDENCE_REPORT.md

chronometric-emergence/frontier-evidence-v2/README.md

@@ -0,0 +1,31 @@
+# Chronometric emergence — frontier evidence repair v2
+
+**Author: Angus Muffatti**
+
+This is the adversarial repair layer created after an independent audit found that many earlier `PASS` labels proved internal consistency rather than physics. It does not conceal that result. It turns the criticism into executable gates.
+
+## What this package actually does
+
+- independently checks the magnitudes of all eleven three-loop 3PI ledger coefficients by coloured half-edge graph automorphisms;
+- demonstrates that the former v1.3 RG cancellation was imposed rather than derived and formally retracts it;
+- runs a genuine non-Markovian two-time scalar benchmark with independent normal-mode and causal-memory solvers, timestep/memory convergence, commutator, FDT, and energy diagnostics;
+- recalculates the cascade daughter fate and supplies an exact-`Z6` gauge-charged thermalisation repair;
+- evaluates the low-`f_a` ridge for Earth, Sun, white dwarfs, and neutron stars;
+- quantifies a 24-link gauge/deconstruction route to the required UV quality;
+- recomputes the v1.4 LPM diagnostic with the corrected collinear mass combination, executes the v1.5 high-resolution recomputation, and checks weak-scattering scaling;
+- measures the v1.7/v1.8 off-shell model discrepancy and forbids table extrapolation;
+- scans portability and replaces the old binary `PASS` vocabulary with an evidence taxonomy.
+
+## Run
+
+```bash
+python -m pip install -e '.[test]'
+python -m frontier_evidence.run_all --repo-root ../..
+pytest -q
+```
+
+The run writes machine-readable evidence to `results/` and a consolidated report to `docs/FRONTIER_EVIDENCE_REPORT.md`.
+
+## Honest status
+
+The package earns a **credible evidence frontier** for the narrow formal theorem, one-loop threshold interpretation, code-quality repairs, and scalar two-time infrastructure. It does **not** yet earn the claimed non-Abelian/cosmological frontier: neutron-star conversion, full UV completion, complete hard+LPM matching, and a real 3PI evolution engine remain load-bearing.

chronometric-emergence/frontier-evidence-v2/pyproject.toml

@@ -0,0 +1,29 @@
+[build-system]
+requires = ["setuptools>=69", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "chronometric-frontier-evidence"
+version = "2.0.0"
+description = "Adversarial evidence and repair suite for chronometric emergence"
+requires-python = ">=3.11"
+dependencies = [
+  "numpy>=1.26",
+  "scipy>=1.12",
+  "sympy>=1.12",
+  "networkx>=3.2",
+  "PyYAML>=6.0",
+]
+
+[project.optional-dependencies]
+test = ["pytest>=8.0"]
+
+[tool.setuptools]
+package-dir = {"" = "src"}
+
+[tool.setuptools.packages.find]
+where = ["src"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-ra"

chronometric-emergence/frontier-evidence-v2/src/frontier_evidence/__init__.py

@@ -0,0 +1,4 @@
+"""Evidence-repair suite for the chronometric-emergence programme."""
+
+__all__ = ["__version__"]
+__version__ = "2.0.0"

chronometric-emergence/frontier-evidence-v2/src/frontier_evidence/compact_objects.py

@@ -0,0 +1,110 @@
+"""Compact-object audit of the low-f_a chronometric ridge."""
+from __future__ import annotations
+
+from dataclasses import dataclass, asdict
+import json
+from pathlib import Path
+
+
+MPL = 2.435e18
+P_QCD = 2.0 / 27.0
+Q_CONVERSION = 0.633135
+
+
+@dataclass(frozen=True)
+class RidgePoint:
+    f_a_gev: float = 2.435e10
+    epsilon: float = 2.70e-13
+    d_g: float = 1.0e-6
+    eta_heavy_max: float = 612.0
+
+
+OBJECTS = {
+    "Earth": 6.95e-10,
+    "Sun_surface": 2.12e-6,
+    "white_dwarf_typical": 1.0e-4,
+    "neutron_star_low_compactness": 0.12,
+    "neutron_star_typical": 0.20,
+    "neutron_star_high_compactness": 0.30,
+}
+
+
+def scalar_compactness(phi_newton: float, point: RidgePoint) -> float:
+    return 2.0 * P_QCD * point.epsilon * (MPL / point.f_a_gev) ** 2 * phi_newton
+
+
+def run(point: RidgePoint | None = None) -> dict:
+    p = point or RidgePoint()
+    objects = []
+    for name, potential in OBJECTS.items():
+        q = scalar_compactness(potential, p)
+        objects.append(
+            {
+                "object": name,
+                "newtonian_potential": potential,
+                "q": q,
+                "q_over_conversion_threshold": q / Q_CONVERSION,
+                "conversion_necessary_condition_met": q > Q_CONVERSION,
+            }
+        )
+
+    y = p.f_a_gev / MPL
+    epsilon_from_ridge = 27.0 * p.d_g * y
+    relation_error = abs(epsilon_from_ridge - p.epsilon) / p.epsilon
+    safe_y_ns = 4.0 * p.d_g * OBJECTS["neutron_star_typical"] / Q_CONVERSION
+    safe_fa_ns = safe_y_ns * MPL
+    eta_at_no_conversion_boundary = 6.12 * p.d_g / safe_y_ns
+    q_eta_ratio_ns = scalar_compactness(OBJECTS["neutron_star_typical"], p) / p.eta_heavy_max
+
+    blocker = any(
+        row["conversion_necessary_condition_met"]
+        for row in objects
+        if row["object"].startswith("neutron_star")
+    )
+    return {
+        "evidence_class": "PREDICTION",
+        "point": asdict(p),
+        "ridge_relation_epsilon_27dg_y_relative_error": relation_error,
+        "objects": objects,
+        "no_conversion_fa_min_GeV_for_typical_NS_at_same_dg": safe_fa_ns,
+        "maximum_heavy_focusing_at_NS_no_conversion_boundary": eta_at_no_conversion_boundary,
+        "q_over_eta_for_typical_NS": q_eta_ratio_ns,
+        "single_source_ridge_no_go": {
+            "statement": (
+                "Within the archived N=6 ridge relations, the same d_g/f_a combination "
+                "controls heavy-threshold focusing and compact-object conversion. Requiring "
+                "q_NS<q_conv caps the heavy focusing measure near 4.8, below the demonstrated "
+                "strong-attractor regime."
+            ),
+            "assumptions": [
+                "one QCD-coupled ratio mode",
+                "epsilon=27 d_g f_a/M_P",
+                "archived heavy-threshold focusing fit",
+                "spherical conversion capacity criterion",
+                "no additional density-dependent screening or transient selector bath",
+            ],
+        },
+        "neutron_star_blocker": blocker,
+        "scientific_status": "BLOCKED" if blocker else "CONDITIONAL",
+        "allowed_repairs": [
+            "derive and confront a converted/scalarised neutron-star branch with binary-pulsar data",
+            "introduce a demonstrated compact-object screening mechanism",
+            "separate the transient cosmological focusing interaction from the late QCD coupling",
+        ],
+    }
+
+
+def write_results(output_dir: Path) -> dict:
+    output_dir.mkdir(parents=True, exist_ok=True)
+    result = run()
+    (output_dir / "compact_object_results.json").write_text(json.dumps(result, indent=2) + "\n", encoding="utf-8")
+    return result
+
+
+if __name__ == "__main__":
+    import argparse
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--output-dir", type=Path, default=Path("results"))
+    args = parser.parse_args()
+    answer = write_results(args.output_dir)
+    print(json.dumps(answer, indent=2))