[amaybaum]

SUBSTATE-1 — observer-relative finiteness and native observation

Current head: 77482a8c43d0570661694bd572849d36602ab73a. Research draft, unmerged.

Status: label maintenance commits 0982e04aee4519659655582cd81e4829ea4df87d, f9347916addaa649a3eb366a69226e5a11d127cd and 77482a8c43d0570661694bd572849d36602ab73a by writer-substate-labels (#1031 c.6676219154). The independent review of 0982e04aee4519659655582cd81e4829ea4df87d returned ACCEPT WITH CHANGES (c.6676347136); the delta check of f9347916addaa649a3eb366a69226e5a11d127cd returned ACCEPT (c.6676394785); 77482a8c43d0570661694bd572849d36602ab73a applies its remaining status-line finding; independent delta check pending; merge not authorized.

The work separates exact finite-horizon representation, observable memory/revival, and physical availability of measurement operations. Manuscripts and publication theorem grades are unchanged.

Current status

• Last review of the head: REVISE. The NC-1 exact-head re-review of d4a80fc8 found a conceptual-definition problem (c.6555867706).
• NC-1 is paused. The owner paused NC-1 at d4a80fc8 (c.6556077148). The definition review round closed as DEFINITION-OPEN, with Q1 and Q7 open (c.6559049909).
• Formal hold retired (c.6605353090). Priority, readiness and authorization are recorded on SUBSTATE-1: observer-relative finiteness without a finite-total-carrier requirement #1007.
• Units after the table below:
  • 54ee14db was accepted at c981a006 (c.6542967608);
  • the later-access test returned LA-EQUAL / EMPTY at 8af2d904 (c.6543875618).
• Acceptance labels. The "Independent ACCEPT" labels in this description report verdicts recorded on this PR. They were written by model agents and posted from the amaybaum account; none is human.
  • On main, verification/programmes/substratum/continuous-substates/README.md:13–17 (at f051916a) classifies them as author-reported and not independent verification.
  • The same README, at :22–31, lists the paper proofs as not independently reviewed on main.
  • See six-PR audit finding 5 (#1031 c.6665380893). Its required repair and completion criteria are at #1007 c.6666811169.

Research checkpoints

Checkpoint	Result and scope
3cb09a15	A/B ACCEPT: scoped recurrence/gauge/measurement conclusions; retained infinite-tape XOR counterexample with common hidden prior, all-time P-divisibility and positive three-time conditional memory. XOR and all-L Fourier arguments remain paper proofs. CI 41118151283 passed.
7c4e45f7	Independent ACCEPT: side-three scalar torus wave candidate has Γ1=J2, Γ2=swap a.e. for an added distributed readout, while its global update has infinite order. This excludes a common global finite-order return, not individual recurrence. Physical readout/preparation/settings remain unsourced.
917d8992	Conditional ACCEPT: 78-tick q10→port5 first-arrival reduction. Under the explicit original-time uniform-key law and prospective record-horizon declaration, a fresh bank key masks the raw symbol. No full transcript/carrier independence follows.
58004838	Independent ACCEPT: exact conditional record/key posterior and full global two-slice pushforward, with terminal branches retained. Public key recovery requires an inverse constant on the public-record fibre. This reduces the joint-kernel problem; it does not evaluate a native operational carrier or supply a setting pair.

| edf8aa44 | Native actuation diagnostic; independent review pending. Fixed transported raw phase at L+79 (sufficient slice radii77/78), exact local input-translation identities, and actual-table residual-field analysis. Normal/optimized bounded checks pass. No accessible decoder, actuator, reusable boundary or two-setting instrument is supplied. |

| f24e1784 | Autonomous two-tick q10 feedback test; independent review pending. Actual sign returns to A with a12-bit mask. A free original key permutes the returned current triple; its uniform conditional law rejects a clean raw echo under the named prior. Previous-slice/full-state correlations remain, and encoded/distributed gates are not excluded. |

| f17fe14c | Independent ACCEPT for the scoped internal packet: exact42-bit sign parity; uniform64 same-site pair under the explicit joint-key prior; original-key counterexample to deterministic raw-packet closure. Bounded independent controls pass. Stochastic operational closure and a physical gate remain open. |

| 923be23b | Independent ACCEPT for the exact conditional packet/public-record kernel over two passive ticks L+3→L+4→L+5. Original joint product priors, triangular stage maps, correlated environment and terminal limits verified. No readable packet, selectable measurement or arbitrary reuse is established. |

| 5865311c | Independent scoped ACCEPT with a nonblocking arithmetic-control finding: full-G proof accepted, but old factor Counter marginalized G.B/D. Physical programmable measurement remains unestablished. |

| 53bf7ac8 | Independent ACCEPT for the minimal full-G test repair. All G fields are retained; the previously escaping output/G.B-D correlation and record-erasure mutants reject. The scoped selector conclusion is unchanged; no programmable native measurement is established. |

| 4baaca43 | Independent scientific ACCEPT for the bounded apparatus-register construction attempt, with a minor comment accuracy nit corrected by529bc5be. A previous-shell A register gives an exact two-tick packet-boundary D flip with all residual fields retained. Pullback forces an original pinned-cell change, so the clean one-register pair fails the fixed preparation. No selectable apparatus or reusable measurement is established. |

| 529bc5be | Independent exact-head ACCEPT, no remaining findings. Minimal comment-only repair of the apparatus inverse-control description. No executable, proof, preparation, access or scientific conclusion changed. |

| 89cbceea | Independent ACCEPT, no remaining findings. Only the original programme in the retained fixed512-controller plus affine-E/N class preserves the complete old phase and whole q10 triple context by context. No second matched-input setting or operational apparatus is established. |

| 4632c3c7 | Independent ACCEPT, no remaining findings. Existing source cluster computes J.B=1 XOR a*mR in two ticks from affine payloads, with A/D residuals retained. The tested programme changes Q on all64 source rows, so delivery, matched full input and reusable measurement remain unestablished. |

| 6aa0ed03 | Independent ACCEPT, no blocking findings. All16 paired affine source programmes preserve the local nonlinear computation but fail matched incoming-carrier preparation. A shared reflection-plane counterexample explains the obstruction. No measurement outputs or reuse are tested. |

| 94ad6e50 | Independent ACCEPT, no blocking findings. Sixteen independent constant-offset pairs preserve the complete local computation; two asymmetric pairs escape the balanced common witness, but change Q on 40/48 and 43/48 continuing contexts. Both fail complete matched input; terminal diagnostics remain. No measurement output or reuse is tested. |

| 32d1ce0f | Independent ACCEPT, no blocking findings. Known-seed affine K offsets, then existing-E compensation: exact finite-domain elimination returns UNSAT in the fixed N/J/G geometry. A continuing context has 12 local/old-record-compatible pairs, each requiring different E for the two unknown inputs. No coefficient, complete-carrier or measurement acceptance is inferred. |

| fbbdb606 | Independent ACCEPT, no blocking findings. One additional existing J source enlarges the local/old-record domain, but the fixed G/N plus J/K/E family remains UNSAT. A continuing-context certificate exhausts14 source triples, each requiring different E values for the two unknown inputs. No measurement output or reuse is tested. |

| f1a6fa25 | Independent ACCEPT, no blocking findings. Targeted G/N screen: the whole earlier q8 record fixes G; N admits four common controls at k0 but has an empty continuing domain at k1. Global affine synthesis is rejected before coefficient coupling. A causal paper argument extends this scoped obstruction to simultaneous G/N coordination with these source positions. No full-carrier or measurement acceptance is inferred. |

| 0085faeb | Independent ACCEPT, no blocking findings. Causal-interface feasibility map: earlier whole-prefix constraints successively fix Q/C/G; earlier cones lie within the prepared region. A free late-entry coordinate lies outside both packet slices and old public history through L+3, with possible boundary contact at L+4. Physical setting source, public access and original-carrier-compatible computation remain open; no native simulation is run. |

| 488ee2b9 | Independent ACCEPT, no blocking findings. Physical-source audit: record labels cannot prepare the independent private p field under the retained product prior; the original J2/N4 early cone lacks mR; public setting access and outgoing reuse are unconstructed. This is a scoped passive-source/interface barrier, not a universal control no-go. CI41188641449 passed25 jobs with1 intentional skip. No native simulation is run. |

| 6e25dd84 | Independent ACCEPT, no blocking findings. Original-preparation q10 fields contain later mR-dependent nonlinear terms, but all448 tested linear-parity/affine-mask target readouts fail. Eight continuing paired contexts have identical whole q10 triples and different required target bits. A supplied public-port-5 action at S+4 could first contact the packet at L+4; an action delayed until the input cut first arrives at L+80. Native actuator, public setting record and reusable joint output remain unconstructed. Q-only obstruction, not a whole-packet no-go. CI41189377581 passed25 jobs with1 intentional skip. No native simulation is run. |

| 4095d985 | Independent ACCEPT, no blocking findings. Thirteen existing free original triples give a triangular bijection onto the full42-bit packet fibre encoding Q. Under the explicitly retained joint product prior, K has mass8^-13 on each compatible packet (8^-12 after fixing the stipulated V record). The same eight target-sensitive equal-Q pairs have identical full-packet laws. This excludes a packet-only target decoder without separately delivered mR, not environmental or later native computation. Existing public port5 supplies a random bank record, not a compiled setting command. Original benchmark, records/failures and environment correlations are retained. CI41190661465 passed25 jobs with1 intentional skip. |

| 54ee14db | Accepted at c981a006 (c.6542967608): the independent review of 54ee14db returned ACCEPT WITH CHANGES, and the delta re-review of c981a006 returned ACCEPT. Actual S+4 port5 randomized-setting test at K_(L+3), with no write/source added. One42-key original triangular map certifies setting independence at the use cut and identical complete future packet/raw-record kernels at L+4,L+5 under the explicit joint product prior. Raw setting B/D, optional V, failures/timing and global environment correlations are retained. This scoped candidate supplies no distinct declared measurement laws; later/environmental control and other priors/interfaces remain open. |

Evidence boundaries

• Existing Lean CausalReadback.c4r_implies_pIndivisible establishes rooted distinguishability revival → P-indivisibility without hidden-carrier finiteness. The candidate's torus/Haar sourcing proof remains paper-level.
• ObservableMeasureReturn proves the ENNReal null-change iff; the real row-stochastic normalization and composed general measure bridge remain open.
• The finite-horizon S⇔D⇔Q_fb equivalence and finite-carrier conditional recurrence results retain their original scopes.
• Abelian gradient flatness and restricted covariance obstruction are exact; SM §6.5's non-Abelian link candidate and native Yang–Mills action selection remain conditional/open.
• Gauge-pointer and interference/eraser probes are toy/reference controls.
• The scalar torus candidate and nonlinear native-Y construction have different dynamics/priors. They are not a common native realization.

Current kernel and remaining physical gate

The new native randomized-setting test''`` follows the original bank source of the S+4 public port5 record through the unchanged native rule. It is independent of the incoming packet at use, but both setting branches have the same joint future packet/raw-record kernel in the existing L+4,L+5 window under the stated42-key joint product prior. Causal reach is distinguished from a certified nonzero signal and from different operational laws. The full setting triple, optional V, terminal records and conditional global environment remain; equality is not claimed for the entire environment or another future access menu. No new source/write, passive horizon or preparation/access assumption is adopted. Normal/-O results agree. The review found that the probe has no mutation harness, so the mutation-class claim was removed in c981a006. Independent review accepted `c981a006` (c.6542967608). CI41197605842 was queued when it was first recorded here.

The accepted full-packet operand and setting-source unit'''``` checks all14 raw triples under the original product-key law. Its conditional fibre proof shows that enlarging q10 to this packet supplies no extra original-remainder information beyond Q; the required correction still fails on eight paired continuing contexts. The optional whole-V record is retained as a packet coordinate, and privacy is claimed only for port histories through the input cut, not future H=L+79. The original native_route_v13 contract permits arbitrary complements; the result uses the explicitly retained independent-uniform-complement option and is not asserted for arbitrary correlated priors. The physical controller remains uncompiled. Normal/-O checks agree; four mutation classes reject in both modes. Independent ACCEPT and CI41190661465 (25 passed,1 intentional skip) close that exact head. No native field simulation is run.

The accepted later-operand/controller unit''''```` reuses the original native table and preparation. Later mR dependence is present in q10, but this cell alone does not provide the required compensation across all continuing contexts. The exact public-register timing separates an early controller deadline from post-preparation interaction; a hypothetical perturbation is not an available actuator or an operationally distinct instrument. Normal/-O checks agree, with four mutation classes rejected in both modes. Independent ACCEPT and CI41189377581 (25 passed,1 intentional skip) close that exact head. No native field simulation is run.

The accepted physical-source audit'''''````` keeps the late-entry geometry but finds no supplied native p-preparer in the declared passive/read interface. It proves the restricted read-label nonsteering statement under the retained product prior and identifies missing mR in the original J2/N4 early computation. Earliest possible old-port contact is L+83, without a public-decoding/reuse guarantee. Independent ACCEPT and CI41188641449 (25 passed,1 intentional skip) close that exact parent. The #866 definition review ACCEPTs the prospective conditional definition with explicit policy/seed/history conventions; it is not applied or adopted as a native quotient.

The accepted causal-interface map''''''`````` identifies a late coordinate outside both incoming slices and old public history through L+3, with possible next-boundary contact at L+4. Independent exact-head ACCEPT confirms the scoped geometry and prior qualification; CI41186219855 passed25 jobs with1 intentional skip. Physical setting source, access and original-carrier-compatible computation remain unsourced.

The accepted G/N screen'''''''``````` fixes G by q8 and rejects N/J/K/E at a continuing context after its initial four necessary-gate survivors. Independent exact-head ACCEPT confirms both finite certificates and the scoped joint G/N causal implication; CI41183852552 passed25 jobs with1 intentional skip.

The accepted existing-J coordination unit''''''''```````` adds two local/old-record-compatible controls but preserves no matched input in its fixed-G/N family. Independent exact-head ACCEPT verifies the decisive certificate; CI41182673724 passed25 jobs with1 intentional skip.

The accepted known-seed K/E synthesis unit'''''''''````````` returns UNSAT with N/J/G fixed, before affine coupling matters. Independent exact-head ACCEPT reproduced the whole continuing certificate with a separate scalar calculation; CI41181909520 passed25 jobs with1 intentional skip.

The accepted asymmetric delivery unit''''''''''`````````` preserves the complete local computation with unequal constant offsets. Two programmes escape the balanced common witness, but change Q on 40/48 and43/48 continuing source contexts. Independent exact-head ACCEPT confirms the scoped rejection with separate dense/scalar controls; CI41175769613 passed25 jobs with1 intentional skip.

The accepted balanced delivery unit'''''''''''``````````` preserves the complete two-tick J computation across its 16 paired affine source programmes, but none preserves matched input. Reflection-symmetric off-plane signals cancel before the nonlinear map in the relevant central-plane causal cone. Unequal offsets now escape that common witness, while failing the broader source-context input comparison.

The accepted nonlinear-compensation unit''''''''''''```````````` demonstrates a clean local computed B field in the existing J/G/K/N cluster under the full-ball preparation, using only affine candidate payloads. Its complete local triple retains A/D residuals. The computed value is at J,time2 and is not a freely assigned initial N.B; causal timing does not prove delivery. The concrete candidate fails the full incoming-carrier test in all48 continuing source rows. No measurement/output kernel or reuse is established. Independent review accepts this scoped local result and carrier failure.

The accepted coordinated source unit'''''''''''''````````````` keeps the full-ball preparation and tests existing affine controller/source positions. The swapped old-phase controller changes the complete q10 triple encoded in the incoming42-bit packet. A four-context contradiction rules out affine E/N compensation; a rank60 system leaves only the current programme in this fixed class. This is a context-by-context preparation test, not a general native-control no-go or an exclusion of a different operational coarse graining. The required bit product is a necessary subcube constraint, not a sufficient full compensation. The complete output kernel, physical selection and reuse remain open.

The accepted apparatus-register unit''''''''''''''`````````````` starts from accepted baseline53bf7ac8. The proposed pair keeps the entire incoming42-bit packet and current environment fixed, changing only a previous-shell A field. Native evolution gives a genuine clean D-field effect after two ticks, but inverse evolution necessarily changes a source-pinned original cell inside the q10 ready ball. This scoped design is blocked by preparation, not by absence of native coupling. The obstruction uses the retained full-ball preparation; it does not cover a different parity-only preparation. Coordinated original-time apparatus encodings remain open. Full environment correlations and records are retained; no new preparation, access or reset is adopted.

The accepted interface test uses an earlier public selector rather than extending the outgoing passive horizon. It is an entry adapter L+1→L+3 into the accepted successor, not a proof that the old G record is an independent setting of an already prepared L+3 packet. Its two observational selector branches agree on the fixed packet/record transition formula. Global setting/environment correlations remain; no causal do(setting) map or general native-interface no-go is claimed. The older two swapped affine controllers also agree on all64 complete old-q9 target rows, without proving equality of their full outgoing workspace. A setting-dependent native coupling and justified system/apparatus split remain to be supplied.

The accepted conditional calculation keeps both complete slices of q10 and its six neighbors. For each of two passive native ticks, its previous slice is the known current slice, its new center is computed from the packet, and all six new boundary triples are jointly uniform under the explicit original-key product law. Together with eight existing raw public-port increments, each compatible one-step joint output has mass 8^-14. Future-key exclusion gives the same next-step kernel.

The proof uses28 already free/reserved original triples across the two stages (12 complement,16 bank), not a reset or newly inserted noise. Existing off-axis source witnesses avoid the prepared normal-axis directions. The full outgoing environment remains a correlated native pushforward; no independent-bath assertion is made. Causal kernels condition on past records and retain new increments as outputs. Earlier terminal branches have no new read or offered carrier.

The ordinary native next step consumes these microscopic fields without decoding s. This is a bounded internal stochastic description, not an observer-accessible instrument: the packet remains private, no selectable measurement pair or native acceptance predicate is implemented, and arbitrary-depth/composite reuse is not established. The route variant still does not inherit Native A's six-bit pool. The earlier f17fe14 packet result remains pinned to its review; the conditional two-step kernel has now received independent ACCEPT at unchanged923be23b. The next unit must specify a physically permitted control independent of incoming carrier preparation, a complete allowed raw read/flag record, and a usable outgoing native boundary. Another passive extension is not the next research target.

Verification and governance

Normal/optimized bounded controls pass; independent review includes wrong-mode, wrong-time and terminal-read mutations for the respective checkpoints. These controls are not full native astronomical-scale replays or new Lean certificates.

Accepted selector baseline53bf7ac8 has successful ordinary PR CI 41163524102 and independent repair ACCEPT, with its coverage finding resolved. Apparatus head4baaca43 has successful ordinary CI 41167056415:25 passed,1 skipped, and passing normal/optimized bounded controls. The independent review identified a minor inverse-test comment error;529bc5be corrects only those two comment lines. Independent review accepts the scientific result and exact repaired head, with the nit resolved and no remaining findings. Repair CI 41167642332 completed successfully at529bc5be:25 passed,1 skipped. Accepted coordinated-source head89cbceea has independent ACCEPT with no remaining findings and successful automatic CI 41168570027:25 passed,1 intentionally skipped. Normal/optimized author and independent controls pass. No correction or head movement was needed for that review. Accepted nonlinear head4632c3c7 has independent ACCEPT, no remaining findings, and successful automatic CI 41170543474:25 passed,1 intentionally skipped. Normal/optimized author and independent controls pass; no correction or head movement was needed for that review. Accepted balanced-interface head6aa0ed03 has independent ACCEPT, no blocking findings, and successful ordinary CI 41174180421:25 passed,1 intentionally skipped. No correction or head movement was needed. Accepted asymmetric head94ad6e50 has passing normal/optimized author controls and eight rejected mutation classes in both modes; independent acceptance is recorded with no blocking findings; automatic CI 41175769613 passed on this exact head: 25 jobs succeeded, 1 intentionally skipped. Accepted synthesis head32d1ce0f has identical normal/-O author results and seven rejected mutation classes in both modes; independent review accepted this exact head and CI41181909520 passed:25 jobs,1 intentionally skipped. Accepted J-coordination headfbbdb606 has identical normal/-O author results and six rejected mutation classes in both modes; independent review accepted this exact head and CI41182673724 passed:25 jobs,1 intentionally skipped. Accepted G/N headf1a6fa25 has identical normal/-O author results and seven rejected mutation classes in both modes. Independent review accepted this exact head; CI41183852552 passed:25 jobs,1 intentionally skipped. Accepted causal-map head0085faeb has identical normal/-O geometry results and four rejected mutation classes in both modes. Independent exact-head review accepted the scoped geometry; CI41186219855 passed:25 jobs,1 intentionally skipped. Accepted physical-source head488ee2b9 has identical normal/-O exact source/probability results and four rejected mutation classes in both modes. Independent review accepted this exact head; CI41188641449 passed:25 jobs,1 intentionally skipped. Accepted later-operand head6e25dd84 has identical normal/-O exact table/algebra/geometry results and four rejected mutation classes in both modes. Independent review accepted this exact head; CI41189377581 passed:25 jobs,1 intentionally skipped. Accepted packet-operand head4095d985 has identical normal/-O exact geometry/conditional-fibre controls and four rejected mutation classes in both modes. Independent review accepted this exact head; CI41190661465 passed:25 jobs,1 intentionally skipped. The randomized-setting unit 54ee14db has identical normal/-O actual-source geometry and full-joint comparison controls. Its mutation-class claim was removed in c981a006, and independent review accepted c981a006 (c.6542967608). CI41197605842 was queued when it was first recorded here. Earlier acceptance and CI are not transferred. Ordinary PR CI tests a synthetic merge and is not an F/E dispatch attestation.

Shared KTOKEN maintenance #1018 is on main; frozen controls/receipts remain unchanged. No F/E designation, special workflow, readiness, research merge or manuscript promotion follows from this PR.

---

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 d4a80fc871473285a7c848a4984940d52817ff8c and the existing work, reviews and evidence on this branch. This comment changes nothing on the branch.
• Obligation row: #1074, row SUBSTATE-1: observer-relative finiteness without a finite-total-carrier requirement #1007 (Observer-relative finiteness, SUBSTATE-1). This PR's NC-1 definition is also recorded in row P0: physical observable carrier and evolution selection #866, O-SET (Setting-dependent interaction, NC-1). Priority, prerequisites, readiness and authorization for this PR's work are recorded only in those rows.
• 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]

Independent review of the label-maintenance commit 0982e04a: ACCEPT WITH CHANGES

The commit was authorized by #1031 c.6676219154 and written by writer-substate-labels.

1. Structure. Single parent d4a80fc8. One file changed, continuous-substates/README.md (+27/−3). Only label and status text changed; no proof, statement, Lean, census or probe. The change is additive: no rebase, no force-push, and the old head remains an ancestor.

2. Inserted block. Branch README:7–26 is byte-identical to main README:13–32 at 92f4f270.

   • Every reference inside the block resolves on the branch.
   • All 30 probes on the branch are byte-identical to main.
   • Leaving out main's "Standing of this directory" paragraph is correct.

3. Review sentences. These are accurate. c.6529351978 was posted by the PR author. Its heading reads "REVISE (text, controls and one split; no mathematical errors found)", and it names no reviewers.

4. Required change: CI-status descriptions that main has corrected.

   • Run 41094894776 is a pull_request run, which tests a synthetic merge. README:288, which this commit edited, and equivalence-and-d3-scope.md:46–47 call it an "exact-head" run.
   • The branch's modules are byte-identical to the versions landed on main at 82f0e119 (run 41871661965, green). They differ from a02e197d only in doc comments (798558b7).

   Make one additive commit using main's text:

   • README:204–207: replace with main README:210–220, minus its last sentence about the main-only doc-comment edit.
   • README:288: replace with main README:301–307, verbatim.
   • observable-recurrence.md:5–6: "were kernel-built on a02e197d." → "are kernel-built in the Mathlib bridge job."
   • equivalence-and-d3-scope.md:46–47: use main's wording ("are both kernel-built, with their #print axioms reports checked, by the Mathlib bridge job (Build; release-gate lean-axioms).").
   • Leave ObservableMeasureReturn.lean:22–23 unchanged. It is byte-identical to main's landed file, and editing it would be a Lean change.

5. Items the writer left alone are sound.

   • The native-*.md headers are byte-identical to main, which covers them through the README block.
   • The NC-1 preregistration's "accepted" and "independent review" uses are plans or references to Research (WIP): SUBSTATE-1 — internal sub-states and visible recurrence #1009 units, not verdicts.
   • No other stale label was found.

6. NC-1 outcomes. No file contradicts REVISE (c.6555867706) or DEFINITION-OPEN (c.6559049909). These outcomes are recorded in the PR body.

7. Register. Clean.

---

Generated by Claude Code

[amaybaum]

Delta check of f9347916: ACCEPT

1. Additive and in scope. The commit has a single parent, 0982e04a, and changes 3 files (+22/−9): the README Observable paragraph (:204–213) and GradientLinkFlatness sentence (:292–298), observable-recurrence.md:5–6, and equivalence-and-d3-scope.md:46–48. No Lean, census, probe or proof file is touched.

2. Copied verbatim from main. Both notes files are now byte-identical to main at 92f4f270. The README differs from main only in two places:

   • main's "Standing of this directory" paragraph, correctly left out;
   • main's doc-comment sentence, discussed in item 4.

3. The copied statements are true for the branch.

   • The module blobs equal those landed at 82f0e119: 49b2ae37, 2f608c21, 141aa9fe.
   • 82f0e119 is the evidence(§A.42, landing 3 of 3): land #1009's three SUBSTATE-1 Lean modules #1100 merge that landed them.
   • In run 41871661965 (push, 82f0e119), the "Build" and "Release gate" steps both succeeded; the release gate includes lean-axioms.
   • Run 41094894776 is a pull_request run.
   • Commit 798558b changes only comments.

4. One remaining inaccurate status line. ObservableMeasureReturn.lean:22–23 on the branch reads "Status: kernel-built on exact Research (WIP): SUBSTATE-1 — internal sub-states and visible recurrence #1009 head a02e197d (automatic CI 41094894776 …)". That text was written in 798558b, after a02e197d. The run it names is a pull_request run, not an exact-head run. It also contradicts the corrected README.

   main fixed this in e0476bf with a two-line doc-comment edit. The proposed fix is a third additive commit that:

   • applies that edit verbatim;
   • restores the README sentence "The doc-comment edit after 82f0e119 changes no declaration."

   The module would then be byte-identical to main (e94fa68b), with no declaration change.

5. Register. Clean.

Coordinator decision. The line is a status description, and the authorization (#1031 c.6676219154) covers reconciling stale status descriptions. The edit is comment-only and makes the file byte-identical to main's, so it is applied as a third maintenance commit, followed by a delta check. If the owner reads "administrative" as excluding any change to a .lean file, even comment-only, this commit can be reverted additively.

---

Generated by Claude Code

[amaybaum]

Delta check of 77482a8c: ACCEPT, by mechanical check.

• 77482a8c has a single parent, f9347916, and changes 2 files (+4/−3).
• ObservableMeasureReturn.lean: only the two "Status:" doc-comment lines (ll.22–23), inside the /-! … -/ block. This is main's e0476bf edit applied verbatim, and no declaration changes. The resulting blob is e94fa68b389641063109a0e11583b8870c42a114, identical to main at 92f4f270.
• README: restores main's sentence "The doc-comment edit after 82f0e119 changes no declaration.". Branch ll.204–214 now equal main README:210–220.
• What remains different from main: the branch's maintenance-touched files match main except for main's "Standing of this directory" paragraph, which is left out on purpose.
• Research (WIP): SUBSTATE-1 — internal sub-states and visible recurrence #1009 label maintenance is complete. Across the three commits 0982e04a, f9347916 and 77482a8c:
  • the old head d4a80fc8 is an ancestor;
  • NC-1's REVISE and DEFINITION-OPEN are preserved;
  • no proof, statement, declaration or census entry changed.

Merge is not authorized.

---

Generated by Claude Code

[amaybaum]

Owner ruling on the ObservableMeasureReturn.lean documentation correction. amaybaum, Claude Code session chat, 2026-11-06; recorded verbatim at #1031 c.6678263940:

#1009: Keep the ObservableMeasureReturn.lean documentation correction at 77482a8c. I verified that it copies the two status-comment lines from main, leaves declarations unchanged, and passed its mechanical delta check. Reverting would restore an inaccurate CI description. The PR remains draft and unmerged; this is not scientific acceptance of NC-1.

No revert will be made. The PR stays a draft and unmerged.

---

Generated by Claude Code