Shay ad37d03b19 docs(adr-0244,adr-0245): D4 Phase 0 — governance reconciliation

Reconciles ADR-0244 against its own D0 blockers and commits ADR-0245 as
a real companion ADR (previously a dangling reference in ADR-0244 §4).

ADR-0244 governance annotation expanded (11 items, all body text below
preserved verbatim for R&D provenance; annotation governs on conflict):
  1. Q_top proof (already landed) now explicitly drops the §2.2 egress
     ΔQ_top conjunct (it is a hollow, always-true discriminator).
  2. §4-vs-§2 contradiction RESOLVED — new §4a is the governing spec;
     §4 kept verbatim as the (contradictory) original R&D sketch.
  3. §2.1 axis construction was underspecified (packs ship dim-3
     directions, not 32-vectors) — resolved via a grade-1 lift at the
     e1/e2/e3 slots (algebra.cl41.basis_vector), not embed_point.
  4. §2.2 amended: signed overlap (never abs()'d — anti-alignment must
     stay distinguishable from orthogonality); leakage norm is the
     positive-definite Euclidean coefficient norm, not the indefinite
     Cl(4,1) inner product (which (+,+,+,+,-) permits to vanish for
     nonzero leakage).
  5. Inalienability reframed as 5 testable layers (algebraic / runtime
     / pipeline / operational / semantic) — only the first is a math
     guarantee.
  6. Paraphrase-invariance reworded as an empirical, measured property;
     Q_top conservation scoped to pure versors only (relax_to_ground
     can converge to non-versor superpositions — the ADR-0243 SD-A
     subtlety).
  7. boundary_ids activation (currently stored, never evaluated).
  8. Identity manifold is FROZEN — biography holonomy accumulates
     separately and never mutates the identity subspace.
  9-11. Filename correction (multimodal_lifecycle.py doesn't exist;
     real file is cognitive_lifecycle.py), ADR-0245 cross-link,
     theological citation labeled ESV + analogy disclaimer.

New §4a supersedes §4 with a typed, Gram/leakage-correct specification
(ManifoldConditioningError, IdentityGateRefusal, bounded/abstaining
C_id, isfinite + shape guards — no bare assert) for Phase 1/2 to build
against under TDD.

ADR-0245 committed Proposed, verbatim from the R&D export, with a
governance annotation mapping its 4 decisions + Sec 3 acceptance gate
against what the cohesion-directive arc already landed (Rust f32 GP,
_cached_eigh, digest hot-path all done; f64->f32 cast + Sec 3's
speedup/collision proofs open — tracked as D4 Phase 4/5).

Audit doc (adr-0244-cohesion-directive-audit-2026-07-17.md) wrinkles
1-2 and Q5 sequencing updated from 'pending'/'must reconcile' to
resolved, cross-referencing this landing.

[Verification]: in-worktree smoke gate 176 passed; provenance/ADR
governance pins (test_adr_0241_governance_p12.py,
test_adr_0242_topological_quarantine.py, test_third_door_cohesion.py)
30 passed — confirms ADR-0244 stays Proposed (no status flip) and no
existing ADR provenance guard was disturbed. Docs-only change, no
executable surface touched.

2026-07-17 15:32:09 -07:00

9.3 KiB

Raw Blame History

Cohesion Directive Audit — repo state vs. `engineering_cohesion_refactoring_directive.md`

Date: 2026-07-17 Scope: Code-level verification of the seven mandates in docs/analysis/engineering_cohesion_refactoring_directive.md against main @ 24078b11 (post ADR-0243 Phase 3, which landed after the directive was drafted). Related: ADR-0244, ADR-0241, ADR-0242, ADR-0243, docs/analysis/core_cohesion_master_plan.md.

Verdict

The directive is directionally sound but was drafted against a pre-Phase-3 snapshot and, in places, against imagined code. Two mandates (2, 4+5) are accurate and actionable as written. Mandate 1's diagnosis is right but its prescription is already built — the real gap is different (f64, not f32). Mandate 6 is ~85% already implemented; one legibility residual remains. Mandates 3 and 7 describe seams that do not exist yet — they are ADR-0244 implementation work, not refactors of drift. Notable: the default=str + 24-char-digest drift the directive flags was replicated into Lane C (biography_wiring.py) after the directive was written — the drift is spreading, raising the priority of the semantic-rigor batch.

Per-mandate findings

#	Mandate	Verdict	Evidence
1	Rust `geometric_product` fast path	Diagnosis right, prescription stale	The 32×32 CPython loop exists (`algebra/cl41.py:116`). But no `_rust_cl41` module — `algebra/backend.py:67` already implements the exact f32-gated Rust dispatch the directive prescribes (opt-in via `CORE_BACKEND=rust`). The real gap: the hot wave-field substrate runs f64; Rust `geometric_product_f64` exists (`core-rs/src/cl41.rs:126`) but is not exported in `lib.rs`. The D9 parity suite explicitly pins against f32-truncating f64 workloads — the directive's f32-only gate never fires on the actual hot path.
2	`_cached_eigh` memoization	Accurate — apply as written	`relax_to_ground` runs fresh `np.linalg.eigh` per call on dense H (`cognitive_lifecycle.py:546`); `ProblemHamiltonian` is frozen + content-addressed — an ideal memoization key.
3	f64→f32 serving cast	Premise false today; prospective	`identity.py` has no numpy/f32 path — pure-Python scalar heuristics; no ψ handoff to `IdentityCheck` exists. The seam this mandate casts across only comes into being with ADR-0244 §2.1–2.2. `cognitive_lifecycle` is also A-04-banned from serve. → fold into the ADR-0244 arc (D4).
4	Full 256-bit digests; kill `default=str`	Accurate; 3 live sites + spreading	`cognitive_lifecycle.py:128–133` (`default=str` + `[:24]`, incl. `_psi_digest`), `self_authorship.py:42–43`, and `biography_wiring.py:125–126` (Lane C, post-directive). `holographic_vault.py:74` is already full-length; `TurnEvent.trace_hash` is already full SHA-256 (the directive's trace-hash claim is stale — already satisfied). CRDT-collision rationale overstates current exposure (multi-writer deferred), but the fix is right and cheap.
5	Little-endian byte-order contract	Accurate; zero-risk	No byte-order coercion at any digest site. `astype('<f8')` is a byte-level no-op on our LE platforms → digests unchanged, contract made explicit. Pure hardening.
6	"No sampled unimodality violation"	~85% already built (ADR-0242 V1 + Lane A)	`fibonacci_search.py` already has the sampled check, typed `OptimizationFailure`, fail-closed nonfinite/bounds, the bracketed-interval contract, and "never a bare float." Lane A's `kappa_calibration.py` is proposal-only and R-04-compliant. Residual: `propose_kappa_from_search` returns baseline κ=1.0 on failure — but κ=1.0 is the "parameters unchanged" no-op the directive demands, and the typed failure is already the second tuple element. The fix is legibility + the reason rename, not a behavior change.
7	Insertion-order-independent allocation	Premise stale; no live consumer	`atlas_packing.py` is already ordinal-reconstructible (`ALLOCATOR_VERSION` pinned, "layout regenerates from identity + ordinal k") and has zero production call sites. Building `AnchorAllocator` now would be a hollow gate; its consumer arrives with ADR-0244 §2.9. → fold into the ADR-0244 arc (D4).

Spec-level wrinkles (all surfaced)

ADR-0244 §2.3 Q_top — RESOLVED, proven vacuous — central I₅ in odd Cl(4,1) ⇒ the charge collapses identically to 0 on every versor state (the #19 pseudoscalar failure mode), confirmed empirically (evals/adr_0244_qtop_vacuity). See Q4 below. Retired from egress; ADR-0244 annotated.
ADR-0244 §4 contradicted §2.1–2.2 — RESOLVED at D4 Phase 0 — per-axis resonance vs. Gram-projection/leakage-norm/ManifoldConditioningError; the then-dangling "ADR-0245" reference is now a real, committed companion ADR; the bare assert byte-order guard is replaced by typed ValueError guards (incl. isfinite). §2 governs; ADR-0244 §4a is the reconciled specification (see ADR-0244 governance annotation items 2–4, 10 and docs/handoff/ADR-0244-D4-IMPLEMENTATION-PLAN.md Phase 0).
Directive §4 says "six criteria" but lists five — the missing one is the Mechanical-Sympathy gate (see below); criteria 1–2 cover Semantic Rigor, 3–4 honesty/falsifiability, 5 Autonomy/Third-Door — Pillar I had no acceptance criterion.
Quarantine wording — "every refactored module must reside strictly inside evals/" can't apply to an in-place algebra/cl41.py refactor. Adopted reading: new capabilities live in evals/; in-place core refactors keep the A-04 transitive serve-quarantine + smoke/fast-lane gates until acceptance.
Doc placement — the directive's canonical path is docs/analysis/; it sat untracked at docs/. The untracked docs/research/core_cohesion_master_plan.md and docs/research/ADR-0243-…md are stale raw re-exports (the research master-plan predates the AGENTS.md R-01 doctrine note the tracked docs/analysis/ copy carries); removed, not committed.
Directive perf numbers (~40µs/call GP, 50–200µs eigh) are unverified assertions — benchmarks/apple_uma_mechanical_sympathy.py measures them as the D2 evidence rather than taking them on faith.

Amended mandate → work-item mapping

Directive mandate	Where it lands	Amendment
M4 + M5 (digests, byte-order)	D1 semantic rigor	apply as written; +Lane C site
M2 (`_cached_eigh`)	D2 mechanical sympathy	apply as written
M1 (Rust GP)	D2 mechanical sympathy	f64, bit-identical parity (not the already-built f32 gate)
M6 (unimodality)	D3 search honesty	legibility + rename; behavior already correct
M3 (f64→f32 cast)	D4 ADR-0244 arc	seam does not exist yet
M7 (allocator)	D4 ADR-0244 arc	no live consumer yet

The sixth acceptance criterion (Pillar I coverage gap)

Directive §4 lists five gates but claims six. The missing one closes the Mechanical-Sympathy pillar, which otherwise has no acceptance test:

6. Mechanical-Sympathy Gain. An accelerated path must demonstrate a measured, reproducible M1/UMA speedup (committed benchmark artifact, benchmarks/apple_uma_mechanical_sympathy.py) while the parity gate (criterion 1) holds. An "accelerated" path that does not beat the pure-Python baseline is rejected. This prevents shipping a fast path that isn't faster.

Decisions (per the "most genius/optimal" directive)

Q1 (M1): expose Rust f64 GP; parity contract = bit-identical (matched i-major scatter order, no FMA), not tol-matched — else flipping CORE_BACKEND=rust breaks I-02 replay-determinism (_psi_digest/wave-residual bytes shift). Fail-closed on any ULP divergence over N=10,000; if bit-identity is unattainable, the f64 Rust path is not shipped.
Q2 (M6): κ=1.0-on-failure is the required "parameters unchanged" no-op, and the typed OptimizationFailure is already returned alongside — so the fix is legibility (a test pinning that failure surfaces as OptimizationFailure) + the reason rename to sampled_unimodality_violation_observed, not breaking the working seam.
Q3: the sixth criterion is the Mechanical-Sympathy gate above; adopt the quarantine reading in wrinkle #4.
Q4 (ADR-0244 §2.3): RESOLVED — Q_top proven vacuous, retire from egress. The non-vacuity demonstration was attempted and failed: evals/adr_0244_qtop_vacuity (pinned by tests/test_adr_0244_qtop_vacuity.py) shows Q_top = 0.000e+00 exactly on every versor, redundant with the closure residual off the manifold, and blind to a versor-preserving identity attack (ΔQ_top = 0 passes an attack that drops identity overlap to 0.963). Do not wire it as an egress gate; keep only as a closure-derived diagnostic.
Q5 sequencing: ADR-0243 Phase 4 → Phase 5 → D0 (this) → D1 (semantic rigor) → D2 (mechanical sympathy) → D3 (search honesty). D4 = ADR-0244 + ADR-0245 implementation — IN PROGRESS. Both prerequisites (§4-vs-§2 reconciliation, Q_top vacuity) are resolved in the ADRs as of D4 Phase 0 (2026-07-17); full phased plan + live progress: docs/handoff/ADR-0244-D4-IMPLEMENTATION-PLAN.md.

Every D-batch PR: in-worktree smoke gate + fast lane green before merge; merge-commit on explicit authorization; local-first CI.

9.3 KiB Raw Blame History Unescape Escape

Cohesion Directive Audit — repo state vs. engineering_cohesion_refactoring_directive.md