[amaybaum]

This is the guiding draft PR for research unit MR-1, measurement as physical record formation. It is coordinated through #866, with #914 rows 20 (native instrument), 14 ((Obs, μ)) and 12/L3 (wave-operator lift).

Status:

• Preregistration frozen at 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c, by owner designation in comment 6549225579.
• Stage 1 authorized: literature-first and paper-only, under comment 6549252522.
• Current state: stage 1 is stopped at the literature gate, as recorded in comment 6549290592.
  • The first-hand reads of Catani et al., PRL 129, 240401 (2022) and Mazurek et al. 2016 are complete. Both have their gate MET, with scope caveats.
  • Kochen–Specker 1967 has its gate NOT MET because of access: only pp. 59–68 of 59–87 were available, and §5 was not read.
  • K★, which includes the CC2 instance, is blocked until the owner decides.
• Stage-1 commits: 7f1b1f39 adds amendments/stage1-log.json (the literature log). CI run 41243523916 passed.

Frozen file: verification/programmes/oi-qm/reconstruction/measurement-record-formation/preregistration.md, 770 lines. The drafting snapshot D is main 5914c17c.

Review history: four independent rounds plus an exact-diff check (comments 6549054511 and 6549175444):

Commit	Verdict
377d2ee3	REVISE
6e6f7cbb	REVISE
32d0d299	ACCEPT WITH CHANGES
77ebafa9	ACCEPT FOR FREEZE-READINESS
1a236f33	one-line fix, CONFIRM

CI was green at the freeze (run 41241775407).

Contents in brief:

• §3: a record-forming interaction (S, φ, μ, π, Z, Acc, ρ, Out, Res). Absorption, loss of interference and irreversibility are kept separate.
• §4 A/B/C, on paper:
  • every record is a function of the complete state;
  • (C) must give up determinism, records-as-configuration or states-as-points;
  • a C-vs-A separation holds only relative to a frozen K that OI holds, and at D only K1 qualifies;
  • (C) is not adopted.
• §5 tests:
  • T0 tests coherence before marking. The other tests cover graded marking, repeat, undo, eraser (full masses), composition, two systems, contextuality, source deletion and inverse audit.
  • Preparation equivalence follows P0: physical observable carrier and evolution selection #866.
  • The duality witness is V + P ≤ 1, using carrier predictability.
  • CC1–CC8 are the countercontrols.
• §6 tokens: there are three separate axes. "Nonclassical" requires outcome separation from CC1–CC7, with a mandatory CC8-unseparated sub-label. A scoped GAP is a valid result, and there is no global no-go.

Not authorized:

• Stage 2 or 3 tests;
• any computation, simulation or experiment;
• Python or Lean;
• premise adoption, including any non-functional record primitive;
• altering the preregistration or its criteria;
• F/E or F2 promotion;
• manuscript, ROADMAP or OI ↔ QM equivalence: proof gate (dependency index; carries #885) #914 edits;
• dispatch;
• merge.

The PR stays draft and unmerged.

🤖 Generated with Claude Code

https://claude.ai/code/session_013T6pfKuv5mi3MXtKjH2tZb

[amaybaum]

MR-1 preregistration: independent review record

Verdict: ACCEPT FOR FREEZE-READINESS at 77ebafa9720b9e4b4e89b00d092523e58937ed26, from an independent read-only reviewer that is not the writer. CI at that commit is run 41241370016, still pending.

Review rounds (each commit is a single-parent child of the one before; no amend or force-push):

Commit	Verdict	Main points
377d2ee3	REVISE, 3 blocking	false complementarity claim (noncontextual models cannot reproduce the duality relation; Catani et al., PRA 108, 022207); self-referential witness W3; the NONCLASSICAL/W1 token did not separate the Kochen–Specker countercontrol
6e6f7cbb	REVISE, 2 blocking	duality bound applied to marker distinguishability instead of carrier predictability; countercontrol separation measured against the benchmark instead of the candidate's own law
32d0d299	ACCEPT WITH CHANGES	the CC2 instance constraint must be a cited OI holding; the proof of Props. V.1–V.2 is in an unread PRL; W2-with-signalling row
77ebafa9	ACCEPT FOR FREEZE-READINESS	one wording nit, below

Owner's four conditions, as the reviewer answered them for 77ebafa9:

• (a) Duality criterion: yes. V + P ≤ 1 uses the carrier's predictability P, on orbit-realizable preparations certified as §5.3 operational equivalences. It claims no bound on V + D.
• (b) External benchmark for Q1: yes. π and μ range over admissible classes, and bounds must be cited OI holdings. At D, only K1 is usable.
• (c) Seven-control requirement: yes.
  • CC1–CC7 must be outcome-separated from the candidate's own law, and that is kept distinct from control separation against the benchmark.
  • CC8-unseparated is a mandatory sub-label.
  • "Nonclassical" is relative to the frozen CC2 instance; the class-level ceiling is CONTEXTUAL with CC2-unseparated.
• (d) Literature status and execution blocker: yes. Three load-bearing sources are not yet read first-hand: Mazurek et al. 2016; the Kochen–Specker instance; and Catani et al., PRL 129, 240401 (arXiv:2207.11779). No execution step depending on a claim from any of them may proceed until it is read first-hand and recorded.

Outstanding nit (L747): the blocker sentence says "a claim from either", but it now covers three sources. It should read "from any of them". This is not fixed at 77ebafa9.

Not done: no computation of any physical, candidate, control or algebraic quantity; no Lean; no candidate; no premise adoption; no manuscript/ROADMAP/#914 change; no F designation, dispatch or merge. Freeze and stage-1 execution await separate owner decisions.

---

Generated by Claude Code

[amaybaum]

Owner decision — MR-1 preregistration: Option 1 (2026-10-31)

APPROVED: one-line blocker wording correction before freeze. I verified #1023 at exact head 77ebafa9720b9e4b4e89b00d092523e58937ed26, the independent ACCEPT FOR FREEZE-READINESS review, and §9 line 747. Replace only “a claim from either” with “a claim from any of them” in the execution-blocker sentence. The three sources meant by this restriction are Mazurek et al. (2016), the Kochen–Specker source/instance, and Catani et al. (2022 PRL, arXiv:2207.11779). Do not narrow or otherwise amend the restriction.

Exact sequence:

1. The designated writer makes a minimal single-line commit on the existing MR-1 guiding draft branch, with no other edits to scientific definitions, predicates, outcomes, countercontrols CC1–CC8, benchmarks, assumptions or reference status.
2. An independent reviewer checks the exact diff and confirms that the change is purely clarifying and that the preregistration's prior substantive review acceptance still holds at the new head.
3. Verify ordinary draft-PR CI on the new exact commit. Only after that verification and reviewer confirmation should the new SHA be designated as the preregistration freeze. The old SHA 77ebafa9 is review-accepted but not designated as frozen by this decision.
4. Retain all four review rounds and their corrections in the PR record.

Execution blocker retained: no stage-1 step that depends on a claim from any of those three sources may proceed until that source is read first-hand and the reading/provenance recorded. In particular, an abstract-only reading does not satisfy this gate. No candidate screening or computation has been authorized here.

Scope: This is authorization to correct the wording and obtain independent exact-head review/CI only. No MR-1 Stage 1 candidate-freeze execution, experiments, Python/Lean computations, physical assumption adoption, changes to manuscripts/ROADMAP/#914 grades, special workflow dispatch, F/E designation or merge. MR-1 stays draft and unmerged. Any additional substantive edit returns for review rather than being included silently.

[amaybaum]

MR-1 preregistration: freeze candidate ready for owner designation

This completes the owner's Option 1 sequence (comment 6549092945).

1. One-line correction at 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c, a single-parent child of 77ebafa9. The diff is exactly one line in §9 (L747): "a claim from either" → "a claim from any of them". No other text, whitespace or file changed.
2. Independent exact-diff review: CONFIRM.
   • The change is purely clarifying. The restriction still names all three sources: Mazurek et al. 2016; the Kochen–Specker instance; Catani et al., PRL 129, 240401 (arXiv:2207.11779).
   • It still requires a first-hand read, and an abstract-only read does not count.
   • §7.1 and §8 are unchanged.
   • The earlier ACCEPT FOR FREEZE-READINESS, with (a)–(d) all yes, holds at 1a236f33.
   • Hygiene passes: placement OK, census OK, SC-42 PASS with 0 added-text hits.
3. Ordinary draft-PR CI at 1a236f33: run 41241775407, conclusion success.
4. Review record retained: all four rounds are kept above, in comment 6549054511.

Candidate for owner designation as the MR-1 preregistration freeze: 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c. It has not been designated. 77ebafa9 is review-accepted but is not the freeze.

Execution blocker carried into stage 1: no step that depends on a claim from any of the three sources proceeds until that source is read first-hand and the reading and its provenance are recorded.

Not done or authorized: stage-1 candidate freeze, candidate screening, computation, Python/Lean, premise adoption, manuscript/ROADMAP/#914 change, dispatch, F/E designation, merge. The branch is held at 1a236f33 pending the owner's decision.

---

Generated by Claude Code

[amaybaum]

OWNER DESIGNATION — MR-1 preregistration FROZEN at 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c (2026-10-31)

I hereby designate exact commit 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c as the MR-1 preregistration freeze. This designation follows the approved Option 1 sequence and is for the preregistration only.

Verified before designation:

• Current Research (draft): MR-1 — measurement as physical record formation, preregistration only #1023 guiding draft PR head matches this exact SHA; it remains draft, open and unmerged.
• The only revision from 77ebafa9 is the one-line §9 wording change, “a claim from either” → “a claim from any of them”, without narrowing the restriction.
• Independent exact-diff reviewer CONFIRM at the freeze candidate; the earlier independent ACCEPT FOR FREEZE-READINESS still holds, including all four checks (a) duality criterion, (b) external benchmark, (c) CC1–CC7/CC8 classical-control discrimination and scoped labels, (d) load-bearing literature block.
• Ordinary PR CI run 41241775407 completed success on this exact SHA.
• All four substantive review rounds and the freeze-candidate record are retained in Research (draft): MR-1 — measurement as physical record formation, preregistration only #1023 comments 6549054511 and 6549175444.

What is frozen: verification/programmes/oi-qm/reconstruction/measurement-record-formation/preregistration.md at this exact SHA, including record-forming measurement definition, A/B/C alternatives, full record and outgoing-carrier contract, sequential/undo/eraser/composite tests, classical comparisons, outcome levels, decision function and procedural scope. The established finite-horizon theorem and accepted negative research scopes are unchanged. Do not retroactively adjust the baseline, expected results, controls, hypotheses or verdict criteria.

Unresolved mandatory literature gate: Mazurek et al. (2016), the Kochen–Specker source/instance and Catani et al. (2022 PRL 129, 240401; arXiv:2207.11779) must each be read first-hand in full and their relevant claims/provenance recorded before any stage-1 or later step depending on them. An abstract-only reading is insufficient.

Not authorized by this freeze: Stage 1 candidate selection/freezing, candidate computations, simulations, physical experiments, Lean/Python execution, adoption of non-functional access or a new OI premise, F/E or F2 promotion, manuscripts, ROADMAP/#914 status edits, special workflow dispatch, or merging #1023. Stage 1 requires a separate explicit owner authorization after this freeze. Keep the PR draft and unmerged.

The next work, if separately approved, must first distinguish incomplete access to a larger classical state from any genuinely new physical record-forming primitive. This designation is not evidence that such a primitive exists.

[amaybaum]

Owner decision — AUTHORIZE MR-1 Stage 1 (literature-first, paper-only), 2026-10-31

APPROVED: Stage 1 only, under the frozen MR-1 preregistration at exact commit 1a236f332e8ea12b7fdb125bf8c5098ccd71d36c, designated by owner comment 6549225579. This approval does not authorize Stage 2 tests or Stage 3 instrument/composite execution.

Mandatory order and admissible scope

1. First-hand literature gate first. Read in full, from their primary sources, Mazurek et al. (2016), the specified Kochen–Specker source/instance, and Catani et al., Phys. Rev. Lett. 129, 240401 (2022), arXiv:2207.11779 (including the proof of Props. V.1–V.2). In the append-only Stage 1 provenance log, record exact edition/version/pages, the claim actually consumed, caveats/conditions and any disagreement with the frozen MR-1 text. Abstracts, second-hand summaries and merely citing the paper are insufficient. No dependent decision or candidate screening relying on a claim from these sources before the corresponding first-hand reading is complete and recorded. If a source contradicts a load-bearing frozen assumption, STOP for independent review and an owner decision rather than changing the preregistration.
2. Precommit K★ before any candidate screen, including the independently justified concrete CC2 instance and its OI-held constraint. The source and justification must not be chosen by looking at a candidate's behavior or the desired quantum outcome. Preserve the frozen external benchmark requirement; no bespoke restriction that excludes inconvenient classical models.
3. Perform paper-only structural screening under the frozen MR-1 §3–§7 rules. Identify the system, apparatus, environment, physically implemented setting source, complete accessible record and outgoing carrier, permitted adaptive operations and inverses, ensemble/observation-map status, and sourced dephasing/record process. Distinguish (A) full-state deterministic functional records, (B) stochastic/latent-variable completions, and (C) an actual new primitive. Neither C nor coherence is assumed/adopted by definition.
4. Append-only Stage 1 files amendments/MR-star.md and amendments/stage1-log.json, preserving the required commit order: K★ first, then screen, then either one fully specified 𝔐★ for later evaluation or GAP with the exact unsourced condition and its tested scope. Freeze the entire Stage 2 interface: Acc, test menu/continuations, countercontrols CC1–CC8 and their admissible instantiations, T0–T9, T7 inequalities, source deletion, mutation list, results schema, and paper-derived benchmark formulas. Keep all branch masses, failures, times, references and outgoing states in scope. Do not run those tests.
5. Obtain independent exact-commit review of the Stage 1 amendment, evidence provenance, no-pre-freeze-computation compliance and unchanged preregistration; bring the result back for a separate owner Stage 2 decision. If none qualifies, preserve an Axis 1: GAP scoped to the sources considered, not a universal no-go.

Absolutely no pre-freeze computation: follow frozen §7.1 / clause 0(vi). No Python or Lean, numerical/algebraic candidate calculations, simulation, test execution or candidate/control outcome probes before the reviewed Stage 1 freeze. Only the explicitly allowed repository hygiene operations may execute, with every run logged. Read literature and reason on paper; use neither helper scripts nor quick sanity-check probes. If a calculational necessity or conflict appears, stop and ask rather than silently proceeding. Keep the scientific/writing source separate from any executable experiment.

Not authorized: Stage 2 controls/candidate execution, Stage 3 operational tests, adoption of a non-functional record primitive, altered preregistration or selection criteria, manuscript/ROADMAP/#914 grade changes, F/E or F2 designation, dispatch, merge or alteration of #1009/#1011/PC-1 research records. #1023 remains research-only draft and unmerged. Every new substantive assumption or proposed primitive must be returned to the owner with its exact changed OI premise, independent motivation and falsifiable consequence.

Scientific emphasis: loss of interference alone is a weak discriminator; the critical target is physically sourced coherence before marking and reversible/sequential/composite recombination that cannot be explained by the declared classical controls on the candidate's own full joint records. Quantum-state overlap is not identified with ordinary overlap of classical probability distributions.

[amaybaum]

MR-1 stage 1: Kochen–Specker reading gate not met (access); stopped for owner decision

Under the stage-1 authorization (comment 6549252522), the frozen literature gate requires a first-hand reading in full of Kochen & Specker, J. Math. Mech. 17, 59–87 (1967) before any step that depends on it. The CC2 instance, and therefore K★, depends on it. The gate cannot be met with the access available to this session.

What was obtained (provenance):

• IUMJ archive (iumj.indiana.edu/IUMJ/FTDLOAD/1968/17/17004/pdf): a free 6-page preview, pp. 59–64 only. sha256 d76d792c….
• University-hosted JSTOR download (cqi.inf.usi.ch/qic/Kochen1967.pdf, JSTOR stable 24902153): a cover plus pp. 59–68 only. sha256 fd7af55a….
• Paywalled or restricted: the JSTOR full PDF; the Springer reprint (Hooker ed. 1975, DOI 10.1007/978-94-010-1795-4_17); the archive.org lending copy.

Not read: pp. 69–87. These pages hold the 117-vector construction, the main theorem's exact statement, §4 (the spin-1 realization) and §5 (the classical spin-½ model and the contextual hidden-variable discussion). §5 is the part the CC2 instance needs. The pages that were read (pp. 59–68) contradict nothing in the frozen text, but they are not a reading that satisfies the gate.

Consequences, per the frozen §7.1 and the stage-1 order:

• K★, which includes the CC2 instance, cannot be committed.
• The candidate screen cannot start, because it must follow K★.
• Nothing has been committed for stage 1. The first-hand reads of Mazurek et al. 2016 and Catani et al. PRL 129, 240401 are still in progress and are independent of this.

Options for the owner:

1. Supply a full primary copy, for example through institutional JSTOR or IUMJ access, a library copy of Hooker 1975, or Specker's Selecta (1990). The reading then proceeds as frozen.
2. Authorize a substitute primary source for the CC2 instance. Because §9 names KS 1967, this would be an append-only amendment to the frozen text, reviewed before use. One candidate is Mermin, Rev. Mod. Phys. 65, 803 (1993), which is open on arXiv as 1802.10119 and covers KS and contextual hidden-variable models. These are unread, and none of them satisfies the current gate.
3. Hold stage 1 at the literature gate.

No computation, no candidate and no K★ choice has been made.

---

Generated by Claude Code

[amaybaum]

MR-1 stage 1: owner decision on the Kochen–Specker reading gate (Option 2), recorded from chat

Source: the owner, amaybaum, in the coordinating session's chat on 2026-11-01, before 04:39Z. Asked whether to proceed on that chat message or wait for a GitHub post, the owner chose "Proceed on chat". This comment records that decision. It replies to the stop in comment 6549290592.

Decision (Option 2): authorize a narrowly scoped substitute primary-source route for the Kochen–Specker reading gate, using two complementary sources:

• J. S. Bell, "On the Problem of Hidden Variables in Quantum Mechanics", Rev. Mod. Phys. 38, 447 (1966). It supplies an explicit deterministic hidden-variable construction relevant to CC2.
• N. D. Mermin, "Hidden Variables and the Two Theorems of John Bell", Rev. Mod. Phys. 65, 803 (1993). It covers the contextuality obstruction and its assumptions. Mermin's published corrections must also be considered.

The owner noted that Mermin is not a direct substitute for the original Kochen–Specker spin-½ model. Neither reference automatically supplies the disturbing-marker extension that CC2 requires.

Boundaries set by the owner:

• The substitution is recorded through an append-only, independently reviewed amendment. The frozen commit 1a236f33 stays unchanged.
• Both substitute sources are read first-hand in full. Their exact relevant claims, limitations and provenance are recorded.
• An independently justified concrete CC2 model, with its OI-held constraint, is established before K★ is frozen.
• If that construction cannot be justified, the literature gate stays unresolved. CC2 and the nonclassicality criterion are not weakened.
• There is no candidate screening until the reviewed K★ is frozen, and no computation until the reviewed stage-1 freeze.
• The completed Mazurek and Catani readings remain valid. Research (draft): PC-1 — physical observer–apparatus coupling, preregistration only #1022 is unchanged. Research (draft): MR-1 — measurement as physical record formation, preregistration only #1023 stays draft and unmerged, and there is no stage-2 authorization.

Access confirmed before starting:

• Bell 1966 has full copies at informationphilosopher.com and fisherp.mit.edu.
• Mermin 1993 is arXiv:1802.10119, a posted version that incorporates corrections.
• The INSPIRE record lists two errata: RMP 88, 039902 (2016) and RMP 89, 049901 (2017).

Next:

1. Independent first-hand reads of Bell 1966 and Mermin 1993, including the errata.
2. The substitution amendment, with reading entries appended to stage1-log.json.
3. Independent review.
4. Only then, and only if the CC2 construction can be justified, K★.

---

Generated by Claude Code

[amaybaum]

MR-1 stage 1: K★ blocked (CC2 instance not constructible); reviewed; owner decision needed

Commits on top of frozen 1a236f33, append-only. The preregistration blob is unchanged (c36fb14b).

Commit	Content	Independent review
7f1b1f39	literature log, LIT-1 to LIT-3	ACCEPT WITH CHANGES
da558a8c	KS-substitution amendment (owner Option 2, comment 6553497955); LIT-4 Bell 1966 and LIT-5 Mermin 1993	ACCEPT WITH CHANGES
a9bdeca2	appended corrections REV-1 to REV-7	ACCEPT
3d475935	amendments/MR-star.md §K★: K★ BLOCKED	ACCEPT WITH CHANGES; block judged genuine
41aa6bdd	appended corrections RC-1 to RC-7	ACCEPT

Literature status:

• Catani PRL 2022: MET. Caveat LIT-1-C7: the bound covers finite-Λ models only.
• Mazurek 2016: MET.
• Kochen–Specker 1967: NOT MET; only pp. 59–68 were available.
• Bell 1966: read in full. It gives no disturbing readout, only context dependence, and no marker. Its models are ψ-ontic.
• Mermin 1993: PARTIAL. The text of the errata (RMP 88, 039902; 89, 049901) was not obtained, because APS is blocked from this session. Claims resting on Mermin alone stay blocked: the finite 4-d square, the product rule, and the finite-menu escape claim.

Why K★ is blocked (M1, the sole binding reason):

• At D, the only OI-held constraint is K1. Frozen L561–562 require the constraint to fix the CC2 instance's strength.
• K1 fixes only the native step φ(u_prev, u) = (u, Au − u_prev). The partition, the record map ρ, the marker region, Acc and (π, μ) stay free.
• Nothing OI holds at D selects or independently justifies one instance. Every K1 instance with a readout and a marker is an interface of 𝔐★'s type (frozen L562–563, A1).
• On the K1 route, M2 (disturbing readout) and M3 (disturbing marker) reduce to M1. They are literature gaps only for non-native forms.
• The finite-menu contextuality requirement (c-iv) is the writer's proposed reading, for the owner to accept or reject. It is not frozen text, and the block does not depend on it.
• Three constructions (C-a to C-c) were rejected on sourcing alone. No candidate law was derived, and nothing was computed.

Unchanged: CC2, the nonclassicality criterion, tokens 2.4 and 2.5, and the literature gate. No screen was started.

Owner decision needed. None of these options is adopted.

• D1: supply a selection principle within K1 that fixes one CC2 instance independently of any candidate.
• D2: posit a CC2 form outside K1. This departs from frozen L561–562 and needs a reviewed amendment.
• D3: keep CC2 at class level only, so the instance obligation is removed and 2.4 becomes unreachable. The 2.5 ceiling with CC2-unseparated is already frozen at L563–564. This needs a reviewed amendment.
• Separately: the Mermin errata. Supply their text, allow journals.aps.org/link.aps.org in the environment's network policy, or proceed with the Mermin-only items blocked.

Nothing is authorized beyond stage 1. The PR stays draft and unmerged.

---

Generated by Claude Code

[amaybaum]

MR-1 stage 1: owner decisions on D3 and the Mermin errata, recorded from chat

Source: the owner (amaybaum) gave these decisions in the coordinating session's chat on 2026-11-01, shortly after 05:13Z, replying to comment 6553721062. They are recorded here verbatim in substance.

Decision 1: D3, with CC2 kept at class level only

The review found a genuine obstacle. OI fixes the underlying dynamics but does not determine a unique classical measurement apparatus. Choosing a convenient apparatus just to make the quantum comparison succeed would introduce an unjustified assumption. Under D3:

• the entire admissible classical comparison class is kept;
• no single model is chosen arbitrarily;
• there is no claim that OI has ruled out classical explanations;
• the CONTEXTUAL result category is retained, with CC2-unseparated;
• the stronger NONCLASSICAL category (2.4) is unavailable under the amended protocol.

The owner's framing: "This is a narrowing of what we can claim, not evidence against OI." D3 is to be documented in an append-only, independently reviewed amendment that states explicitly that it changes the previously frozen instance requirement. There is to be no silent reinterpretation of the preregistration, and 1a236f33 stays unchanged.

Decision 2: supply the Mermin errata

The owner verified the original APS PDFs. There are three notices, not two, and all three are to be included in the reading record. The owner's summary:

Notice	Correction
2013 erratum	p. 811: contextual → noncontextual
2016 erratum	Figure 4 caption: three observables in six groups → four observables in five groups
2017 erratum	p. 814: noncontextuality → contextuality

The owner's access does not satisfy the reading gate. The writer must still read the notices first-hand and record their implications.

Next authorized step

1. Prepare the D3 amendment and the errata provenance updates.
2. Obtain independent review and green CI.
3. Then revisit the K★ freeze decision.

Not authorized: stage 2 execution, candidate screening, Lean/Python computation, manuscript changes, or merge.

---

Generated by Claude Code

[amaybaum]

MR-1 stage 1: D3 amendment and Mermin-errata record reviewed and green; K★ freeze decision needed

This implements the owner's decisions in comment 6554185786. All changes are append-only. The preregistration blob at 1a236f33 is unchanged (c36fb14b).

Commit	Content	Independent review
987ecf5e	amendments/D3-CC2-class-level.md (new); KS-substitution.md §8; log records lit5_errata (REV-8) and d3_step	ACCEPT WITH CHANGES
dbf78a86	appended corrections: D3 §D3.9 C-8; MR-star §K★.12; REV-9 and REV-10	ACCEPT

CI run 41279841361 at dbf78a86 concluded success.

D3: what the amendment changes

The amendment states that it changes the frozen instance requirement. It quotes each frozen line it affects:

• L561–564: the CC2-instance obligation is set aside.
• L569, col. 4: no separating test is named for CC2.
• L551–557 and L675–676: these apply to CC2 only. CC2 is not executed, and its separation fields are fixed to "not separated (class level, D3)".
• L529–533 and L620: 2.4 NONCLASSICAL is unreachable; the row text is kept.
• L713: the 2.4 self-test becomes a check that no input yields 2.4.
• L661–664: the commit order no longer places a CC2 instance in K★.
• L45–47 (outcome (2)) and L613 (axis-2 heading): these stand as labels only. Their strongest token is 2.5.

Unchanged: CC1 and CC3–CC8; rows 0 and 2.1–2.9 other than 2.4; T0–T9; the duality and benchmark rules; the K★ timing rule; 0(iii) and 0(vi). The 2.5 sub-labels still name every countercontrol that was not separated. The amendment frames this as a narrowing of claims, not evidence against OI, and states no global no-go. The reviewer found the narrowing hard to vary: re-admitting 2.4 needs a new reviewed amendment with new sources or a selection principle OI holds.

Mermin errata

• Citations: RMP 85, 919 (2013); 88, 039902 (2016); 89, 049901 (2017).
• Access: none was obtained first-hand. APS returned 403, and the domains are outside the environment's network allowlist; every route tried is logged.
• Owner's summary: recorded as owner provenance, not as a first-hand read.
• E2: consistent with the arXiv text.
• E1 and E3: their arXiv locations are the reader's inference, not a mapping.
• Status: LIT-5 stays PARTIAL, and the first-hand read requirement remains open. No step under D3 depends on a Mermin-only claim, as the reviewer verified. The finite square, the product rule and the finite-menu escape were needed only to certify a CC2 instance.

Owner decision needed: freeze K★

Under D3, K★ would be {K1}: the one fixed native step φ(u_prev, u) = (u, Au − u_prev), not a per-law construction, with CC2 at class level and no instance. The decision is whether to designate this as the frozen K★.

Freezing K★ is the precondition for the next authorized stage-1 step: the paper-only structural screen, then one 𝔐★ or a scoped GAP, with the stage-2 interface frozen. The screen does not start before your designation.

Nothing is computed, no candidate has been considered, and nothing is authorized beyond stage 1. The PR stays draft and unmerged.

---

Generated by Claude Code

[amaybaum]

MR-1 stage 1: owner decision to freeze K★, recorded from chat

Source: the owner (amaybaum), in the coordinating session's chat on 2026-11-01 at about 06:25Z, replying to comment 6554387397.

Decision. Once CI is green on the exact commit, and provided the reviewed amendment is unchanged, proceed to the K★ freeze. The owner favors freezing:

• K★ = {K1}: the single native OI dynamics constraint;
• CC2 retained as the full classical comparison class, not an arbitrarily selected model;
• nonclassicality: the stronger 2.4 claim stays unavailable, and the contextuality ceiling stays explicit.

The owner added: "Only after that freeze should the paper-only candidate search begin. The goal is still to find a physical measurement mechanism arising from OI itself, rather than reproduce quantum predictions by construction." No stage-2 authorization or merge is implied.

Conditions checked:

• CI run 41279841361 on dbf78a86 concluded success at 2026-11-01T06:21:38Z.
• The remote branch head is still dbf78a86, the commit the independent reviewer accepted. No change followed the review.

Next:

1. An append-only K★-freeze commit in amendments/MR-star.md and the log, recording K★ = {K1} with CC2 at class level under D3.
2. Independent review and CI.
3. Then the paper-only structural screen under frozen §3–§7 and D3, ending in one fully specified 𝔐★ or a scoped GAP, with the stage-2 interface frozen. Nothing is computed.

---

Generated by Claude Code

[amaybaum]

Formal hold retired (owner decision, chat 2026-11-03). Formal holds are retired as a scheduling mechanism (#1074, rule 7). This PR enters the ordinary backlog.

• Preserved: head af1e9be2ceabd5f6181b8d56b537b7ce1a4a3bba and the existing work, reviews and evidence on this branch. This comment changes nothing on the branch.
• Obligation row: #1074, row P0: physical observable carrier and evolution selection #866, O-REC (Accessible records). Priority, prerequisites, readiness and authorization for this PR's work are recorded only in that row.
• No permission: this change authorizes no work, execution, integration, publication or claim promotion, and waives no review or CI requirement.
• Earlier statements of this PR's hold, on this PR and elsewhere, are superseded by this comment and are not edited. Any other requirement they state, such as a merge gate, stands.

---

Generated by Claude Code

[amaybaum]

Retired without merging (PR-AUDIT-1, #1088). The branch and its commits are kept.

Owner decision (amaybaum, Claude Code session chat, 2026-11-04):

#1023 — MR-1 | Land the five Stage 1 records | Approved as evidence-only integration, followed by handoff to #866 and close without merging.

Closing this PR is administrative. It resolves no scientific question and changes no status.

• Evidence on main: the five stage-1 records landed byte-identical through PR evidence(MR-1, #866): land the five stage-1 records of #1023 #1095, merge commit 8a47132d935c27e69c5571e3c05cb6fef6a1e55d. The scoped GAP is unchanged.
• Accountable home: P0: physical observable carrier and evolution selection #866, row O-REC (handoff). MR-1 stage 2 is not authorized and stays blocked on O-RET and O-ACC.
• Kept:
  • branch research/measurement-record-formation-20261031 at af1e9be2ceabd5f6181b8d56b537b7ce1a4a3bba, and refs/pull/1023/head, including the freeze history the records cite;
  • this PR's discussion.

---

Generated by Claude Code