Phase D: claim ledger and scientific provenance

.github/workflows/ci.yml

@@ -450,16 +450,18 @@ jobs:
 
       - name: Unit tests
         run: |
+          python scripts/validate-claim-ledger.py
           python -m unittest discover -s tests -v
           (cd tests && python test_cosmovirus.py)
           (cd tests && python test_phase_b2.py)
           (cd tests && python test_phase_b3.py)
           (cd tests && python test_phase_b4.py)
           (cd tests && python test_phase_b5.py)
           (cd tests && python test_phase_c.py)
+          (cd tests && python test_phase_d.py)
 
       - name: Static type check
-        run: mypy --strict cosmo_core cosmovirus.py tests/test_cosmovirus.py tests/test_phase_b2.py tests/test_phase_b3.py tests/test_phase_b4.py tests/test_phase_b5.py tests/test_phase_c.py
+        run: mypy --strict cosmo_core cosmovirus.py scripts/validate-claim-ledger.py tests/test_cosmovirus.py tests/test_phase_b2.py tests/test_phase_b3.py tests/test_phase_b4.py tests/test_phase_b5.py tests/test_phase_c.py tests/test_phase_d.py
 
       - name: Bytecode compile check
         run: python -m compileall -q cosmo_core cosmovirus.py tests

CLAIM-LEDGER.md

@@ -0,0 +1,223 @@
+# COSMO Claim Ledger
+
+> Generated canonically from `claims/claim-ledger.json`. Do not hand-edit this file independently.
+
+**Schema:** `COSMO-CLAIMS-D-1`
+
+## Evidence classes
+
+- **FORMAL**
+- **COMPUTATIONAL**
+- **SCIENTIFIC**
+- **HYPOTHESIS**
+- **SYMBOLIC**
+
+## Scientific source registry
+
+### SRC-E8-MATHWORLD
+
+- **Kind:** `scholarly_reference`
+- **Domain:** `mathematics`
+- **Title:** Gosset Polytope — E8 root polytope
+- **Identifiers:** none
+- **Identifier URLs:** none
+- **URL:** https://mathworld.wolfram.com/GossetPolytope.html
+
+### SRC-SPIN8-PTEP-2021
+
+- **Kind:** `scholarly_article`
+- **Domain:** `mathematics`
+- **Title:** Vertex operator superalgebra/sigma model correspondences: The four-torus case
+- **Identifiers:** year=2021
+- **Identifier URLs:** none
+- **URL:** https://academic.oup.com/ptep/article/2021/8/08B102/6353037
+
+### SRC-SIS2-PMID-25590815
+
+- **Kind:** `peer_reviewed_article`
+- **Domain:** `materials_science`
+- **Title:** Two high-pressure phases of SiS2 as missing links between the extremes of only edge-sharing and only corner-sharing tetrahedra
+- **Identifiers:** DOI=10.1021/ic501825r, PMID=25590815
+- **Identifier URLs:** DOI=https://doi.org/10.1021/ic501825r; PMID=https://pubmed.ncbi.nlm.nih.gov/25590815/
+- **URL:** https://pubmed.ncbi.nlm.nih.gov/25590815/
+
+### SRC-HPV16-REFSEQ
+
+- **Kind:** `official_database`
+- **Domain:** `biomedicine`
+- **Title:** Human papillomavirus type 16, complete genome
+- **Identifiers:** RefSeq=NC_001526.4
+- **Identifier URLs:** RefSeq=https://www.ncbi.nlm.nih.gov/nuccore/NC_001526.4
+- **URL:** https://www.ncbi.nlm.nih.gov/nuccore/NC_001526.4
+
+### SRC-HPV16-E6E7-PMID-17645777
+
+- **Kind:** `peer_reviewed_review`
+- **Domain:** `biomedicine`
+- **Title:** Basic mechanisms of high-risk human papillomavirus-induced carcinogenesis: Roles of E6 and E7 proteins
+- **Identifiers:** PMCID=PMC11158331, PMID=17645777
+- **Identifier URLs:** PMCID=https://pmc.ncbi.nlm.nih.gov/articles/PMC11158331/; PMID=https://pubmed.ncbi.nlm.nih.gov/17645777/
+- **URL:** https://pmc.ncbi.nlm.nih.gov/articles/PMC11158331/
+
+### SRC-HPV-P16-PMC8409095
+
+- **Kind:** `peer_reviewed_review`
+- **Domain:** `biomedicine`
+- **Title:** Biology of HPV Mediated Carcinogenesis and Tumor Progression
+- **Identifiers:** PMCID=PMC8409095
+- **Identifier URLs:** PMCID=https://pmc.ncbi.nlm.nih.gov/articles/PMC8409095/
+- **URL:** https://pmc.ncbi.nlm.nih.gov/articles/PMC8409095/
+
+## Claims
+
+### COSMO-D-001 — FORMAL
+
+- **Status:** `SUPPORTED`
+- **Statement:** For every CosmoLayer, six applications of the authoritative Lean transition return the layer to itself.
+- **Sources:** none
+- **Boundary:** This is a theorem about the finite state machine only; it is not a physical periodicity claim.
+- **Repository provenance:**
+  - `cosmovirus.lean` — `theorem six_step_periodic (layer : CosmoLayer) : psiIterate 6 layer = layer := by` (`kernel_checked_theorem`)
+
+### COSMO-D-002 — FORMAL
+
+- **Status:** `SUPPORTED`
+- **Statement:** Every authoritative COSMO layer reaches every other layer within one complete six-state orbit.
+- **Sources:** none
+- **Boundary:** Reachability is defined inside the project state machine.
+- **Repository provenance:**
+  - `cosmovirus.lean` — `theorem every_layer_reachable (source target : CosmoLayer) : ReachesWithinCycle source target := by` (`kernel_checked_theorem`)
+
+### COSMO-D-003 — COMPUTATIONAL
+
+- **Status:** `SUPPORTED`
+- **Statement:** The Phase B3 Python reference generator deterministically constructs 240 unique E8 roots with rank 8 and squared norm 2, and the canonical table is bound to a reviewed SHA-256.
+- **Sources:** SRC-E8-MATHWORLD
+- **Boundary:** This is deterministic computational evidence; the local Lean core does not yet formalize the E8 root construction.
+- **Repository provenance:**
+  - `cosmo_core/e8.py` — `def validate_e8_root_system` (`implementation`)
+  - `tests/test_phase_b3.py` — `test_root_system_report_has_rank_eight_and_norm_two` (`regression`)
+
+### COSMO-D-004 — SCIENTIFIC
+
+- **Status:** `SUPPORTED`
+- **Statement:** Spin(8) has triality symmetry with an S3 outer automorphism action that permutes its vector and two spinor eight-dimensional representations.
+- **Domain:** `mathematics`
+- **Sources:** SRC-SPIN8-PTEP-2021
+- **Boundary:** This mathematical fact does not establish a mechanism connecting Spin(8) triality to HPV, capsids, SiS2, or cosmology.
+- **Repository provenance:**
+  - `cosmovirus.tex` — `COSMO-D-004` (`documented_context`)
+
+### COSMO-D-005 — SCIENTIFIC
+
+- **Status:** `SUPPORTED`
+- **Statement:** The ambient-pressure phase of SiS2 is orthorhombic and contains chains of distorted edge-sharing SiS4 tetrahedra.
+- **Domain:** `materials_science`
+- **Sources:** SRC-SIS2-PMID-25590815
+- **Boundary:** This crystallographic fact does not support COSMO's symbolic substrate or life-code interpretation.
+- **Repository provenance:**
+  - `cosmovirus.tex` — `COSMO-D-005` (`documented_context`)
+
+### COSMO-D-006 — SCIENTIFIC
+
+- **Status:** `SUPPORTED`
+- **Statement:** The NCBI reference sequence used for human papillomavirus type 16 in this ledger is RefSeq NC_001526.4.
+- **Domain:** `biomedicine`
+- **Sources:** SRC-HPV16-REFSEQ
+- **Boundary:** The accession identifies a reference genome; it does not validate COSMO's state-machine use of the HPV16Layer label.
+- **Repository provenance:**
+  - `CLAIM-LEDGER.md` — `NC_001526.4` (`provenance_record`)
+
+### COSMO-D-007 — SCIENTIFIC
+
+- **Status:** `SUPPORTED`
+- **Statement:** High-risk HPV E6 and E7 proteins are established carcinogenesis factors; E6 promotes p53 degradation and E7 disrupts pRb/E2F control.
+- **Domain:** `biomedicine`
+- **Sources:** SRC-HPV16-E6E7-PMID-17645777
+- **Boundary:** This biomedical mechanism is external scientific context and is not a mechanism for COSMO transitions.
+- **Repository provenance:**
+  - `cosmovirus.tex` — `COSMO-D-007` (`documented_context`)
+
+### COSMO-D-008 — SCIENTIFIC
+
+- **Status:** `SUPPORTED_WITH_SCOPE`
+- **Statement:** p16 immunohistochemistry is used as a surrogate marker for HPV-associated disease in some clinical contexts, but p16 positivity is not identical to direct evidence of active E6/E7 transcription.
+- **Domain:** `biomedicine`
+- **Sources:** SRC-HPV-P16-PMC8409095
+- **Boundary:** COSMO must not equate a generic p16-positive label with HPV16 infection or active viral transcription.
+- **Repository provenance:**
+  - `cosmovirus.tex` — `COSMO-D-008` (`documented_context`)
+
+### COSMO-D-009 — SYMBOLIC
+
+- **Status:** `PROJECT_DEFINED`
+- **Statement:** The byte-to-cuneiform labels in COSMO are project-defined symbolic annotations.
+- **Sources:** none
+- **Boundary:** They are not a decipherment, transliteration, translation, archaeological attribution, or historical sentence.
+- **Empirical status:** `NON_EMPIRICAL`
+- **Repository provenance:**
+  - `cosmovirus.lean` — `def cuneiformAnnotation` (`project_definition`)
+  - `KNOWN_LIMITATIONS.md` — `Cuneiform strings are project-defined symbolic annotations` (`scope_boundary`)
+
+### COSMO-D-010 — SYMBOLIC
+
+- **Status:** `PROJECT_DEFINED`
+- **Statement:** COSMO's language of undivided cosmic symmetry, life-code substrate, infected reality, and Ouroboros self-causation is symbolic/interpretive vocabulary.
+- **Sources:** none
+- **Boundary:** These phrases are not empirical cosmology, materials science, virology, or causal-mechanism claims.
+- **Empirical status:** `NON_EMPIRICAL`
+- **Repository provenance:**
+  - `cosmovirus.tex` — `COSMO-D-010` (`symbolic_section`)
+
+### COSMO-D-011 — SYMBOLIC
+
+- **Status:** `PROJECT_DEFINED`
+- **Statement:** The association of Spin(8) triality with HPV capsid branching or trimerization is a symbolic cross-domain association in the current repository.
+- **Sources:** none
+- **Boundary:** The repository does not claim an established biological mechanism connecting Spin(8) triality to HPV capsid assembly.
+- **Empirical status:** `NON_EMPIRICAL`
+- **Repository provenance:**
+  - `cosmovirus.tex` — `capsid branching (COSMO-D-011)` (`historical_symbolic_mapping`)
+
+### COSMO-D-012 — HYPOTHESIS
+
+- **Status:** `PROPOSED`
+- **Statement:** A future quantitatively specified mapping from triality-derived features to an HPV/capsid observable could be tested for predictive value against matched controls.
+- **Sources:** none
+- **Boundary:** No such predictive result is currently claimed.
+- **Falsification protocol:** Pre-register the triality-derived mapping, target observable, dataset split, evaluation metric, matched baselines, and decision threshold before evaluating held-out data.
+- **Rejection condition:** Reject the hypothesis if held-out performance fails the predeclared threshold or is not distinguishable from the matched control baselines.
+- **Controls:** Matched baseline models fixed before held-out evaluation; Held-out data excluded from mapping and threshold selection
+- **Repository provenance:**
+  - `CLAIM-LEDGER.md` — `COSMO-D-012` (`hypothesis_definition`)
+
+### COSMO-D-013 — SYMBOLIC
+
+- **Status:** `PROJECT_DEFINED`
+- **Statement:** The use of SiS2Substrate as a COSMO state name is symbolic project vocabulary rather than evidence that silicon disulfide is a biological life-code substrate.
+- **Sources:** SRC-SIS2-PMID-25590815
+- **Boundary:** The scientific source does not establish a biological substrate role; it supports SiS2 crystal-structure facts only.
+- **Empirical status:** `NON_EMPIRICAL`
+- **Repository provenance:**
+  - `cosmovirus.lean` — `SiS2Substrate` (`state_label`)
+
+### COSMO-D-014 — COMPUTATIONAL
+
+- **Status:** `SUPPORTED`
+- **Statement:** Within the documented SECDED capability, the Phase B5 software storage pipeline deterministically recovers the original TriadicLattice through bytes, ECC, ACGT, corruption, correction, bytes, and cube reconstruction.
+- **Sources:** none
+- **Boundary:** This is a software codec result, not a laboratory DNA-storage or physical Rubik's Cube claim.
+- **Repository provenance:**
+  - `cosmo_core/storage.py` — `def recover_cube_storage` (`implementation`)
+  - `tests/test_phase_b5.py` — `test_full_cube_recovers_one_bit_error_in_every_codeword` (`regression`)
+
+## Governance
+
+1. New cross-domain public-facing claims require a stable `COSMO-D-###` ID.
+2. **SCIENTIFIC** claims require external source records from the same controlled scientific domain.
+3. **HYPOTHESIS** claims must remain `PROPOSED` and include structured falsification criteria.
+4. **SYMBOLIC** claims require `empirical_status = NON_EMPIRICAL`.
+5. **FORMAL** and **COMPUTATIONAL** claims require reviewed, class-appropriate repository provenance.
+6. Multi-accession source records must independently bind every URL-addressable identifier.
+7. Positive public cross-domain causal/mechanistic assertions must carry a ledger claim ID.
+8. This Markdown file must exactly match the canonical JSON rendering.

KNOWN_LIMITATIONS.md

@@ -44,11 +44,16 @@ This file is part of COSMO's trust boundary. A green build means the checked cod
 - MIDI/audio visualization remains outside the deterministic storage core.
 - Experiment manifests and storage receipts bind the exact bytes supplied or recovered. They do not independently authenticate external provenance or scientific validity.
 
-## Scientific interpretation
-
-- Formal verification establishes implications inside the model, not correspondence between model labels and the physical world.
-- Cross-domain E8, quasicrystal, SiS2, HPV16, SEER, and Sumerian relationships require independent domain evidence and, where presented as hypotheses, explicit falsification conditions.
-- The repository's bundled historical source documents include claims and terminology that the September 2026 audits identified as inaccurate or unsupported. They remain archived as project inputs and are not the authority for the corrected computational baseline.
+## Phase D claim/provenance boundary
+
+- `claims/claim-ledger.json` is the Phase D authority for evidence class and provenance; `CLAIM-LEDGER.md` is its reviewed human index. A green ledger validator means the repository satisfies the encoded classification rules, not that every externally sourced claim has been independently replicated by COSMO.
+- **SCIENTIFIC** entries report what cited external sources support. They do not transfer scientific authority to a cross-domain bridge merely because the component domains are individually real.
+- Spin(8) triality is established mathematical context, but the current triality↔HPV capsid mapping is classified **SYMBOLIC**. No biological mechanism is claimed.
+- Ambient-pressure SiS2 structure is externally sourced scientific context; `SiS2Substrate` and “life-code” readings remain **SYMBOLIC** project vocabulary.
+- HPV16 RefSeq `NC_001526.4`, E6/E7 biology, and scoped p16-marker usage are external biomedical context. They do not validate `HPV16Layer` as a biomedical state transition.
+- Cuneiform annotations and cosmological/Ouroboros readings are **SYMBOLIC**, not archaeological decipherment or empirical cosmology.
+- A **HYPOTHESIS** entry must retain explicit falsification criteria. No hypothesis may be promoted to SCIENTIFIC or COMPUTATIONAL merely because a suggestive analogy or numerical match is found.
+- The repository's bundled historical source documents include claims and terminology that the September 2026 audits identified as inaccurate or unsupported. They remain archived as provenance inputs and are not the authority for the corrected ledger.
 
 ## Generated documents
 

README.md

@@ -14,7 +14,7 @@ The repository deliberately separates what is **proved or computed** from what i
 | **HYPOTHESIS** | Testable cross-domain proposal not yet validated | “COSMO hypothesizes …” |
 | **SYMBOLIC** | Artistic, mythological, or interpretive mapping | “COSMO symbolically associates …” |
 
-The current Lean/Python core is intentionally narrow. The six COSMO layer names remain project vocabulary, while the code proves only the finite-state and arithmetic facts explicitly stated in the source.
+The current Lean/Python core is intentionally narrow. The six COSMO layer names remain project vocabulary, while the code proves only the finite-state and arithmetic facts explicitly stated in the source. Phase D makes this separation machine-auditable through [CLAIM-LEDGER.md](CLAIM-LEDGER.md) and `claims/claim-ledger.json`; CI rejects class/provenance drift and requires the Markdown index to match the JSON rendering byte-for-byte.
 
 ## Verified baseline
 
@@ -164,6 +164,31 @@ The Lean core now provides:
 
 These are formal/computational properties of the discrete state machine only. The labels do not make the corresponding scientific or symbolic interpretations true. Category-theory construction remains Phase E work and must derive from this state machine rather than define a second transition semantics.
 
+## Phase D claim ledger and scientific provenance
+
+Phase D turns the evidence classes above into a validated repository contract.
+
+The canonical machine-readable ledger is `claims/claim-ledger.json` with schema `COSMO-CLAIMS-D-1`; [CLAIM-LEDGER.md](CLAIM-LEDGER.md) is the reviewed human index. Stable `COSMO-D-###` identifiers bind claims to:
+
+- one of **FORMAL / COMPUTATIONAL / SCIENTIFIC / HYPOTHESIS / SYMBOLIC**;
+- exact repository provenance anchors;
+- external scholarly/database sources where required;
+- accession/version identifiers where available;
+- an explicit evidence boundary; and
+- mandatory falsification criteria for hypotheses.
+
+`scripts/validate-claim-ledger.py` fails closed if IDs are duplicated/out of order, evidence classes change, provenance anchors disappear, scientific claims lack external sources, hypotheses leave the `PROPOSED` state or lose falsification criteria, symbolic claims lose `empirical_status = NON_EMPIRICAL`, provenance paths escape the repository, or the Markdown rendering diverges from JSON.
+
+The initial ledger records, among other things:
+
+- Phase C six-step closure and reachability as **FORMAL**;
+- the Phase B3 E8 root construction and B5 storage round trip as **COMPUTATIONAL**;
+- Spin(8) triality, SiS2 crystal structure, HPV16 RefSeq **NC_001526.4**, E6/E7 biology, and scoped p16 usage as externally sourced **SCIENTIFIC** context;
+- cuneiform, cosmology/Ouroboros, life-code, and triality↔capsid associations as **SYMBOLIC** unless separately tested; and
+- a triality-to-HPV/capsid predictive bridge only as a **HYPOTHESIS** with a preregistered falsification requirement.
+
+Phase D also rewrites the historical LaTeX so unsupported bridges no longer appear with theorem/scientific force. The creative material is retained, but it is explicitly classified.
+
 ## Repository map
 
 | Path | Purpose |
@@ -176,8 +201,11 @@ These are formal/computational properties of the discrete state machine only. Th
 | `cosmo_core/triadic.py` | Deterministic 8×8×8 ternary lattice, bounded parallel updates, diagnostics, and recovery model |
 | `cosmo_core/storage.py` | B5 cube/bytes/SECDED/ACGT/synthetic-FASTA round-trip and integrity receipts |
 | `cosmo_core/dynamics.py` | Phase C authoritative six-state transition, typed witnesses, paths, and reachability |
+| `claims/claim-ledger.json` | Canonical Phase D machine-readable claim classes, provenance, sources, and falsification boundaries |
+| `CLAIM-LEDGER.md` | Human-readable Phase D claim/source index |
+| `scripts/validate-claim-ledger.py` | CI validator for claim IDs, evidence classes, provenance anchors, sources, and hypothesis boundaries |
 | `cosmovirus.py` | Compatibility-facing Python mirror backed by `cosmo_core` |
-| `tests/` | Python regression tests covering B2-B5 deterministic cores plus Phase C six-state dynamics |
+| `tests/` | Python regression tests covering B2-B5 deterministic cores, Phase C dynamics, and Phase D provenance governance |
 | `scripts/check-lean-trust.sh` | Fast lexical preflight for forbidden Lean source constructs |
 | `scripts/prepare-lean-audit.py` | Frozen manifest, artifact, symlink, path-package, and import-layout validation |
 | `scripts/verify-lean-source-state.py` | Hardened Git dependency-source identity and source-cache receipt verification |
@@ -223,6 +251,8 @@ python -m unittest discover -s tests -v
 (cd tests && python test_phase_b4.py)
 (cd tests && python test_phase_b5.py)
 (cd tests && python test_phase_c.py)
+(cd tests && python test_phase_d.py)
+python scripts/validate-claim-ledger.py
 python -m compileall -q cosmo_core cosmovirus.py tests
 ```
 
@@ -232,7 +262,7 @@ For the CI-equivalent static check:
 
 ```bash
 python -m pip install -r requirements-dev.txt
-mypy --strict cosmo_core cosmovirus.py tests/test_cosmovirus.py tests/test_phase_b2.py tests/test_phase_b3.py tests/test_phase_b4.py tests/test_phase_b5.py tests/test_phase_c.py
+mypy --strict cosmo_core cosmovirus.py scripts/validate-claim-ledger.py tests/test_cosmovirus.py tests/test_phase_b2.py tests/test_phase_b3.py tests/test_phase_b4.py tests/test_phase_b5.py tests/test_phase_c.py tests/test_phase_d.py
 ```
 
 ## Current design status