i'm not like you ▣ THE WORKSHOP

← The Book

08 reference coherence audit

Reference Coherence Audit (Wave 2)

The corpus has been swept against two grammars: a reference grammar (every named entity has a single canonical form; every cross-reference resolves to something extant) and a cosmology grammar (the new experiment-frame canon is uniformly applied; vessel-frame residue has been retired). Findings are organized as an SPPV poset of explicit verification tasks. Each task carries the three SPPV gates: PRE (entry condition), POST (acceptance condition), VERIFY (what was checked).

The verdict is short and load-bearing: the experiment-frame reframe did not survive clean. The world-bible canon (book/meta/02-world-bible-canon.md) was rewritten; the glossary, several codex chapters, and the artifact layer were not. The vessel-frame “surface-deployed operating layer” language survives, by exact phrase, in at least three load-bearing locations. The most consequential survivor is codex/glossary.yaml:812, which is the public-facing API contract for the entire app.

A separate, equally consequential failure: the canon contradicts itself on the Wexler death date (12 Nov 1961 in three canon entries; 26-27 Sep 1961 in one). Every downstream artifact uses 26-27 Sep. The 12 Nov entries are stale.

The third headline: the two-Marisols rule is broken corpus-wide. Canon §10.1 says “they must not be collapsed; flag and check first.” Five book chapters (01, 03, 07, 09, 17) and the convocation minutes treat them as the same person. Story 14 (the technician) extends the collapse further. The breach is not subtle; it is the book’s structural spine.

§1. Audit Plan (SPPV poset)

Coherence Epic — Reference Grammar + Cosmology Coherence
  Feature A: Reference Grammar
    Task A.1: Proper-noun canonical-form registry
    Task A.2: Code-and-case canonical-form registry
    Task A.3: Place canonical-form registry
    Task A.4: Cross-reference resolution
    Task A.5: Glossary ↔ canon coherence
  Feature B: Cosmology Coherence
    Task B.1: Three-layer canon (substrate / operators / observers)
    Task B.2: Experiment-frame survival (vessel-frame retirement)
    Task B.3: Procedure semantics (AM-12.4; owl-as-binding-image)
    Task B.4: Pact terminology
    Task B.5: Operator-caste names
    Task B.6: Compression Event timing
  Feature C: Less-Than-Ideal Implementation
    Task C.1: "I have seen but can't say" anti-patterns
    Task C.2: Load-bearing reliance on contested observations
    Task C.3: Trailing-declarative-sentence tic
    Task C.4: Easter eggs that fail to resolve
    Task C.5: Voice-discipline violations

Each task’s PRE/POST/VERIFY gate is given inline below.

§2. Reference Grammar Findings

A.1 Proper-noun registry

PRE: canon §5 People table at book/meta/02-world-bible-canon.md:137-176; glossary figure entries at codex/glossary.yaml:415-587. POST: every proper noun resolves to one canonical form; alias variation is bounded; canon-disallowed names do not appear in chapters. VERIFY: grep across all chapters + artifacts; cross-check against canon and glossary.

Canonical Variants found Files where each variant appears Issue
L. ▆▆▆▆▆▆▆▆ (Lorenzo) “L.”, “Lorenzo”, “L. ▆▆▆▆▆▆▆▆”, “Renzo” (Bill’s name for him) postcards; microcassette; story 02; ch 16; ch 11; codex ch 2, 4, 7, 15 OK — redaction is 8 chars throughout. “Renzo” is consistently used only by Bill Carlyle.
John Roberts “John”, “John Roberts”, “J. Roberts”, “Roberts”, “Renzo’s John” story 02; story 06; story 07; story 16; canon OK.
William Carlyle “Carlyle”, “Bill Carlyle”, “Bill”, “William Carlyle”, “Lt. Col. (ret.) William Carlyle, USAF” story 09; ch 11; codex ch 2 (FN2), ch 7; postcards; ch 16 OK.
Helen Carlyle “Helen”, “Dr. Helen Carlyle”, “Helen Carlyle” postcards; ch 11; ch 16; canon; glossary MINOR: Helen’s relation to Bill is ambiguous. Canon and glossary both say “niece.” Tonopah notebook (book/chapters/16-tonopah-notebook.md:162) confirms via Bill’s line “Renzo, the niece is the one.” But book/meta/00-book-architecture.md:174 says “Helen as Carlyle’s daughter.” The architect’s intention was daughter; canon overrode to niece. Surviving daughter-language: none in chapters, but the architect doc should be updated to “niece” so it stops being a contradiction-of-record. Also: the book Carlyle chapter 11 line 73 says “our daughter had begun her first year at the law school in Charlottesville” — Bill’s biological daughter is in law school, unnamed; Helen the historian is the niece and a separate person. The corpus survives this reading.
Helen (Hauschild’s wife) “Helen” (in diary), “Helen would have told her” artifacts/letters/02-hauschild-diary-pages.html:423 MINOR: Naming collision with Helen Carlyle. The diary’s “Helen” is Eric Hauschild’s wife, “two years gone” by 2014 (so d. 2012). Distinct person from Helen Carlyle. The collision is recoverable from context but worth flagging — both Helens are connected to dying men and care-recipient narratives, and a careless reader could conflate them. Recommend the artifact be edited to call her “my Helen” or rename to break the collision (one-sentence fix in artifacts/letters/02-hauschild-diary-pages.html).
Daniel Wexler “Wexler”, “Col. Wexler”, “Daniel Wexler”, “Col. Daniel Wexler” canon; document 04; story 09 / ch 11; codex ch 2 (FN2), 7 OK at the name level. CRITICAL date contradiction — see Task A.4 and Task B.6.
Eric Hauschild “Eric Hauschild”, “Eric”, “Marina’s father”, “the dying math teacher”, “the boy from El Paso”, “the Asheville file” ch 13; story 11; codex ch 4 (line 100); artifacts/letters/02-hauschild-diary-pages.html; artifacts/letters/03-pops-notebook-entry-1998.html:267 CRITICAL: Canon §10 Open item #8 says “Story 07’s dying father is canonically unnamed (Marina’s father / Daddy only). The task prompt’s ‘Eric Hauschild’ name was a parent-agent working label; canon does not supply a name. Resist supplying one.” But the name “Eric Hauschild” appears explicitly in: book/chapters/13-chaplains-visit.md:22,50,102,168; artifacts/stories/11-the-chaplains-visit.md:20,48,100,166; codex/chapters/04-maintenance.md:100; artifacts/letters/02-hauschild-diary-pages.html (the entire diary attributes itself to him); artifacts/letters/03-pops-notebook-entry-1998.html:267 (“the Hauschild boy”). The cross-narrative naming is no longer the working label — it is canon-violating, in print, across six artifacts. Either canon §10.8 must be updated to canonize “Eric Hauschild” (architect’s recommendation), or the name must be redacted from the six locations. Cannot remain unresolved.
Ray “Pop” Marquez “Pop”, “Ray”, “Pop Marquez”, “Mr. Marquez”, “Ray Marquez”, “R.M.” story 05 / ch 01; canon; ch 17; ch 16 / story 16 (Lorenzo’s notebook) OK.
Marisol (MRI tech, 38, north of Denver) “Marisol”, “Marisol the MRI tech” story 01; book ch 9; canon §5 line 159 OK as a separate person — but see A.1 line below.
Marisol (Pop’s granddaughter, 11, Phoenix) “Marisol”, “Marisol Marquez” story 05; book ch 1; canon §5 line 160 OK as a separate person — but see A.1 next line.
The Two Marisols collapse “Marisol Marquez” used for both in book chapters 01, 03, 07, 09, 17 book/chapters/01, 03, 07, 09, 17; convocation minutes 3.4 + Roberts marginalia; codex ch 4 line 100 (“Marisol’s autumn calendar entry”) CRITICAL: Canon §10.1 says “Future stories must not collapse them. If a writer wants to canonize ‘story-01 Marisol IS Pop’s granddaughter grown up,’ that is a major canon-level decision; flag and check first.” The book has, in fact, collapsed them — every Marisol chapter in the book uses “Marisol Marquez” — and the convocation minutes’ Roberts marginalia (artifacts/documents/07-convocation-minutes-2026-06-12.html:378) makes it explicit: “Marisol. She is 39 next week. The technician on her file has worked her since ‘98 — Pop’s negotiation”. Canon’s two-different-people rule is dead at the book layer. The corpus has implicitly canonized the architect’s preferred reading (book/meta/00-book-architecture.md item §7.2). Canon §10.1 must be updated, or every book chapter naming her “Marquez” must be edited. Recommend the former — the book has gone too far to reverse, and the architect’s reading is structurally superior. Update canon §10.1 to read “Marisol Marquez at age 11 in Story 05 and Marisol the MRI tech in Story 01 are the same person across a 27-year gap; this was canonized in the book layer.”
Daniel Vasquez “Daniel Vasquez”, “Vasquez”, “V., USN”, “Daniel Vasquez of Lubbock, Texas” ch 13; story 11; ch 16 / story 16 (Lorenzo’s ledger entry as “Subject 9930, name on file Vasquez, V., USN”) OK. Note: canon §5 People table does not include Vasquez. He’s only in glossary as a “placeholder pending canonical expansion” (codex/glossary.yaml:576-579). The chaplain chapter names him fully (“Daniel Vasquez of Lubbock, Texas”). Recommend canon add him; he’s load-bearing for Ch 13 / 15.
Pat Donohoe “Pat”, “Pat Donohoe”, “Donohoe”, “Father” (in dialogue) ch 13; story 11; codex ch 15; book ch 9 (Marisol’s therapist named “Pat”) MAJOR: The “Pat” who is Marisol’s therapist in book ch 9 (Partially-Remembering) (book/chapters/09-partial-remembering.md:94) — “She has, since 2019, been seeing a therapist named Pat” — is unmarked. Pat Donohoe is a Navy chaplain who retired from civilian chaplaincy in Asheville (codex ch 15). A patient in Denver-north-suburbs would not be seeing Asheville chaplain-Pat as therapist. The naming is either (a) Easter egg / wink (the cosmology wants the reader to wonder), or (b) sloppy. If (a), it’s deliberate and should be flagged in canon. If (b), rename. Recommend renaming the therapist in ch 9 (call her “Pat Marrero” or similar) — the structural pull is too strong otherwise.
Maren Sandberg “Maren”, “Sandberg”, “M. Sandberg”, “Maren Sandberg” story 03; document 06; canon; glossary; vignette 03 (her grandfather) OK. Note story 03 introduces her with surname (artifacts/stories/03-the-disinfo-agents-inheritance.md); doc 06 confirms “Sandberg, M.” The architect’s note about surname undisclosure (canon §5 line 170) is from an earlier draft. Surname is in print; canon is accurate.
Joan Whitfield “Joan Whitfield”, “Joan” story 08; book ch 12; canon; glossary OK.
Reema Iyer “Reema”, “Reema Iyer” story 06; book ch 4; glossary OK.
Elin “Elin” (no surname given), “the polar researcher” story 04; book ch 8; canon; glossary OK. Glossary explicitly notes “Elin” — no surname is canonical.
Pratt, Greta “Greta Pratt”, “Pratt”, “GP” (in email avatars) story 03; document 06; canon; glossary OK.
Hartwell, Dale “Dale Hartwell”, “D. Hartwell”, “Hartwell”, “Dale” document 06; canon; glossary OK.
Cole “Cole” story 03 only OK — bounded appearance.
Devraj “Devraj” story 03 only; cited in canon §4 Compression Event lexicon OK.
Ninth officer (working group) “Lt. Col. ▆▆▆▆▆▆, USAF”, “the ninth officer” codex ch 7; book ch 11; canon (“ninth officer never named”) MINOR: Canon does not enumerate the ninth officer’s redaction width. Codex ch 7 line 58 says “six characters” and ch 11 line 20 uses 6 characters in the redaction. L.’s redaction is 8. These are consistent; the corpus is using two distinct redaction-widths to encode two distinct figures. Canon should canonize this (add to §5.10 or §10): “the ninth officer’s redaction is 6 chars; L.’s is 8 chars; never collide.”
Aldiss, Pierce, Mead, Reston, Voss, Whitlock Working-group officers codex ch 7; book ch 11; canon OK on names. The death-year sequence is internally consistent.
Veth-skenn / Sshenn-vu “Veth-skenn”, “Sshenn-vu” story 14; book ch 15; codex ch 2 line 103, ch 4 (multiple); glossary OK at name level. Issue: Sshenn-vu only appears in story 14 / book ch 15 (the technician’s monologue). The glossary calls Sshenn-vu “an operator-class name attested in the cosmology’s deep-time linguistic fragments (placeholder pending canon)” — but he’s used in-text as a senior technician at the technician’s gathering, not as a deep-time fragment. Glossary entry is stale.

Acceptance: A.1 PASSED for most names; FAILED on (a) Eric Hauschild canonicality (canon-violating across 6 artifacts); (b) Marisol Marquez two-person rule (collapsed across 5+ chapters + convocation); (c) Pat therapist-vs-chaplain ambiguity (recommend rename or canonize Easter egg).

A.2 Code-and-case registry

PRE: canon §5 Cases table; glossary case entries. POST: every case identifier appears in canonical format YY-XXX-NNN; every procedure code used is defined. VERIFY: grep all chapters + artifacts for case patterns.

Canonical Variants found Files Issue
CASE 79-BNW-001 “CASE 79-BNW-001”, “79-BNW-001” story 03; document 01; glossary OK.
CASE 81-BNW-014 “CASE 81-BNW-014”, “81-BNW-014” story 03; glossary OK.
CASE 84-MOO-003 “CASE 84-MOO-003” story 03; glossary OK.
CASE 87-AQU-009 “CASE 87-AQU-009” story 03; codex ch 2 line 78 (“Lorenzo’s case ledger includes a CASE 87-AQU-009”); glossary OK.
CASE 88-NMX-014 “CASE 88-NMX-014” microcassette doc 05; glossary OK.
CASE 24-LMA-007 “CASE 24-LMA-007” story 03; document 06; glossary OK.
CASE 61-RR-118 “CASE 61-RR-118” document 01; canon; glossary OK.
CASE 03-RR-441 “CASE 03-RR-441” document 01; glossary OK.
CASE 18-RR-077 “CASE 18-RR-077” document 01; glossary OK.
MAR-26-013 “MAR-26-013” story 03; doc 06; inbox 02; glossary; canon OK; uniformly written.
VK-14-3411-B “VK-14-3411-B” vignette 01; inbox 05; glossary OK.
████-7142 “████-7142”, “▆▆▆▆-7142”, “Subject 7142”, “7142” inbox 03/04; ch 16 / story 16; convocation; glossary OK at the identifier level. Cross-reference inconsistency — see Task A.4 (where is 7142 located?).
AM-12.4 Rev C “AM-12.4 Rev C”, “AM-12.4”, “AM 12.4” every chapter; documents; glossary; canon OK formatting. Stale gloss in glossary — see Task A.5.
PCT-0173 “PCT-0173”, “PCT 0173” (only in glossary alias) extensive OK.
§6 / §3(b) / §11 / §14 / §6 review All inflections extensive OK consistent.
OP-CHANNEL-3 / OP-CHANNEL-7 “OP-CHANNEL-3”, “CHANNEL-7”, “OP-CHANNEL-7” story 02; canon; glossary OK.
P1…P9 All canonical extensive OK.
Cell 4 “Cell 4”, “OPS LIAISON Cell 4” story 02; canon; glossary OK.
Cohort II “Cohort II”, “Cohort 2” (glossary alias only) extensive OK.
Cohort R-7 “Cohort R-7”, “cohort R-7” convocation 3.4; architect doc MINOR: “Cohort R-7” appears in the architect doc and convocation minutes but is not defined in canon or glossary. Canon §5 cases table doesn’t list cohorts. Glossary has no entry for it. Either define it or remove the usage. (Canon §3 mentions “the seven-year cohort” and “slow cohort” and “accelerated cohort” but no R-7.)
Sentinel Cartograph LLC “Sentinel Cartograph”, “Sentinel”, “Sentinel Cartograph LLC” extensive OK.
Logistics Support Group “Logistics Support Group”, “LSG”, “Logistics Support Group, LLC.” story 02; story 06; canon; glossary OK.
HEARTH-1 / ALPINE-A / REDWOOD-3 / BRIDGE-7 / EMBER-0 satiation venues doc 03 (calendar); canon; glossary OK.
LANTERN-9 “LANTERN-9” (rotation handler venue) canon §5 line 153 only MINOR: LANTERN-9 appears once, in canon, referring to L.’s rotation venue. Not referenced anywhere else. Either remove from canon or seed at least one artifact.

Acceptance: A.2 PASSED except (a) Cohort R-7 undefined; (b) LANTERN-9 dangling.

A.3 Place registry

PRE: canon §6 Geography table. POST: every place uses canonical geographic dimensions and characterizations. VERIFY: grep place names + dimensions.

Canonical Variants found Files Issue
Lake Vostok “Vostok”, “Lake Vostok”, “Vostok station” extensive OK.
Vostok Bell anomaly: 105 × 75 km “105 by 75 kilometers”, “105 × 75 km”, “approximately 105 kilometers by 75 kilometers” codex ch 1, ch 3; canon; glossary OK, uniform.
Mariana / Challenger Deep depth: 10,935 m “10,935 meters”, “10,935 m” codex ch 1 line 40; glossary; canon OK, uniform. (Note: real-world recent re-measurements give 10,925 ± 4 m or 10,994 m via different methods. The corpus settles on 10,935 m. Internally consistent.)
Kola Superdeep Borehole: 12,262 m “12,262 m”, “12,262 meters” codex ch 1; glossary; canon OK.
Putorana Plateau “800 × 500 km”, “Eight hundred kilometers by five hundred” codex ch 1; canon; glossary OK.
Tibesti Mountains: hot springs 37°C “37 °C”, “thirty-seven degrees Celsius”, “37 degrees” codex ch 1; canon; glossary OK.
Atlantic abyssal plain “Atlantic abyssal”, “Atlantic abyssal plain” extensive OK.
CCZ (Clarion-Clipperton Zone) “Clarion-Clipperton Zone”, “CCZ” codex ch 1, ch 3; canon; glossary OK.
Mizpah Hotel (Tonopah, NV) “Mizpah”, “Mizpah Hotel”, “the Mizpah” postcards; ch 16 / story 16; glossary OK.
Tonopah “Tonopah”, “Tonopah NV” extensive OK.
Marfa “Marfa”, “Marfa lights”, “Marfa, Texas” extensive OK.
Truth or Consequences, NM “T or C”, “Truth or Consequences”, “T-or-C” postcards; ch 16 / story 16 OK.
False Cape, Virginia “False Cape”, “False Cape VA”, “False Cape preserve” story 09 / ch 11; doc 04; canon OK on place. Date contradiction under Task B.6.
Wright-Patterson AFB “Wright-Patterson AFB”, “WPAFB”, “Wright-Patterson” extensive OK.
Reston “Reston”, “Reston/Dulles”, “Reston Building C” extensive OK.
Holloman AFB “Holloman”, “Holloman AFB”, “Holloman Golf Classic” story 05; canon OK.
Asheville, NC “Asheville”, “west Asheville” ch 13; codex ch 15; canon OK.
Alamogordo, NM “Alamogordo” story 05; ch 1; ch 17 OK.
Berkeley (Marina’s home — referenced as such in canon §5 line 157) canon MINOR: Marina is “50s, Berkeley” per canon, but in book ch 2 (Clear Air) and story 07 Marina is at Eric’s bedside in NC, then “flies back to California” per ch 13 line 170 (“before she flew back to California”). Geography is canonical but worth confirming Berkeley is still her base — the chapters don’t say Berkeley by name. Recommend one line in ch 02 to anchor “Berkeley.”

Acceptance: A.3 PASSED. Geographic data is uniform across the corpus.

A.4 Cross-reference resolution

PRE: every “see Ch X” / “per story Y” / “Lorenzo’s notebook entry N” must resolve. POST: no dangling references. VERIFY: grep cross-references; verify each target exists.

Cross-reference Source Target exists? Issue
“Lorenzo’s October 2019 entries” codex ch 1 line 108 YES — artifacts/stories/16-the-tonopah-notebook.md and book/chapters/16-tonopah-notebook.md OK.
“story 01: Marisol’s 2003 overlay binds 83 minutes to ‘an owl on a fencepost’” canon §3 line 102 YES — story 01 / ch 9 OK.
“Story 02 / story 06 / story 07” cross-refs canon §5 People; codex ch 2 (FN2); doc 06 YES, all extant OK.
“Vignette 01 (Galina, the 2014 Core)” architect doc; ch 8 inset YES — artifacts/vignettes/01-the-2014-core.md OK.
“Vignette 02 (Lakeside, second visit)” architect; book ch 11.5 in architecture YES — artifacts/vignettes/02-lakeside-second-visit.md OK.
“Vignette 03 (the handler, Maren’s grandfather)” architect; canon §5 line 170 YES — artifacts/vignettes/03-the-handler.md OK.
“Carlyle memoir, Chapter 4” codex ch 2 FN2 PARTIAL — Carlyle’s memoir is presented as having “chapters” but the artifact (artifacts/stories/09-carlyle-memoir.md / book/chapters/11-norfolk-briefing.md) is one chapter, not a multi-chapter memoir. FN2 says “Chapter 4” of the memoir; this is a self-consistent in-universe reference (the memoir has internal structure that the corpus presents only one chapter of). OK if intended that way; flag if the reader could be confused. MINOR.
“Subject 7142, registry entry 1994” (convocation 3.4) vs “Subject 7142, Marisol Marquez” (ch 16) vs Subject 7142’s NE Sector primary-care intake (inbox 04) three artifacts PARTIAL — Subject 7142 in inbox 04 is geographically “NE Sector”; in ch 16 / story 16 and the convocation Roberts marginalia, 7142 is Marisol Marquez (Pop’s negotiation, southwest US); inbox 04 has the recall onset in May 2026 with a primary-care presentation in NE Sector. Marisol the MRI tech is “north of Denver” — Denver is not “NE Sector” unless the apparatus uses an unusual sector grid. MAJOR: the convocation locks in 7142 = Marisol Marquez; inbox 04 locks in 7142 = NE Sector PCP intake. Reconcile or break. Recommend the convocation reference to “Pop’s negotiation” be re-keyed to a different subject number (e.g., 7143) and 7142 be left as the NE Sector intake from inbox 04.
“CASE 87-AQU-009 in Maren’s grandfather’s box” (canon §5) vs “Lorenzo’s case ledger includes a CASE 87-AQU-009” (codex ch 2 line 78) two sources MINOR — same case # appears in both Maren’s grandfather’s box and Lorenzo’s ledger. Possibly intentional (case is jointly tracked across two desks). Worth canonizing one way.
“the convocation minutes” (book ch 14) book ch 14 → artifacts/documents/07-convocation-minutes-2026-06-12.html YES OK.
“the Bennewitz anniversary archive sweep” inbox 05; architect doc EXISTS as concept; no archive document. MINOR: referenced but not pointed at any concrete artifact. OK as flavor; not load-bearing.
“the 1986 acoustic annex” / “Quantico GSA cage” canon; story 03; doc 06 EXISTS as concept; no annex document. OK.
“the AARI’s 2017 ‘Anomalous returns at the eastern margin, 1998–2014’ that Elin has on her hard drive” canon §1 line 32; codex ch 3 line 28 EXISTS as in-universe document; no artifact realizes it. OK.
“the dictabelt of the 26 September briefing” doc 04 (1962 Wexler memo) EXISTS as in-universe object; doc 05 (1989 dictabelt transcript) is a different dictabelt. OK.
“the 1989 dictabelt (microcassette) transcript” architect; canon §5 (L.); doc 05 YES OK.
“Pop’s small black notebooks” story 05 / ch 1; architect; ch 17 (“the small black notebooks gone, somewhere, she didn’t know where”) PARTIAL — artifacts/letters/03-pops-notebook-entry-1998.html exists as one notebook entry. The architect’s “fourteen notebooks” reading list note is unrealized. The careful reader can locate one entry; the architect’s “Marquez, R. — field-narrative, 1979–2007, partial set, 14 notebooks (donated)” line in book ch 12 is canonized in the architect doc but not in the convocation minutes themselves. Recommend the convocation minutes include this reading-list entry, or drop the architect promise.
“Pat’s footlocker / yellow legal pad with 22 sentences” ch 13 / story 11 EXISTS as in-universe object; no artifact realizes it. OK as character interiority.
“Joan Whitfield’s MacDowell manuscript” book ch 12; canon §10 #5 EXISTS as in-universe; story 08 / ch 12 is the chapter that gestures at it. OK.
“the 1965 Klepac correspondence” / “1967 Mosquito-Range four-light sighting” canon; codex ch 3; clipping 01 YES — artifacts/clippings/01-leadville-herald-1967.html is the 1967 clipping. The 1965 correspondence is referenced but not realized. OK as concept.
“the Hauschild teaching-journal pages (1965, 1972, 1979, 1986, 1993, 2000, 2007, 2014, 2021)” codex ch 4 line 100 YES — artifacts/letters/02-hauschild-diary-pages.html realizes some of these. But codex says cycles 1965–2021 (9 cycles); diary realizes ~4 cycle dates. Not a dangling ref so long as it’s “pages” not “all pages.” OK.
“the 1981 Roswell-road letter” canon §3 line 109; architect ch 3 YES — artifacts/letters/01-roswell-road-letter-1981.html OK.
Convocation chapter cross-references to “Pop” as author of “field-narrative annexes” in the reading list architect promise (book/meta/00-book-architecture.md:301) NOT FULFILLED in the actual convocation minutes (artifacts/documents/07-convocation-minutes-2026-06-12.html). The architect promised Pop in the reading list; the minutes don’t carry it. MINOR: a missed easter-egg payoff (see Task C.4).
Convocation chapter cross-references to “Helen Carlyle” as named source of postcards architect promise (architecture doc) The convocation does discuss the Wexler papers / Carlyle estate; mentions “Helen Carlyle” — partial fulfillment. OK.
Convocation references to “Joan named in case-carry margin annotations” architect promise NOT FULFILLED — Joan is not named in the convocation minutes. MINOR: a missed payoff.
Architect’s plan for “the cumbia” callback in convocation observer margin architect promise (book/meta/00-book-architecture.md:323) NOT FULFILLED — Roberts’ marginalia mentions Pop’s negotiation but does not call back the cumbia. MINOR: deliberate or missed; recommend the architect’s cumbia line in the closing margin annotation, since the spine resolves there.

Acceptance: A.4 mostly PASSED. FAILED on Subject 7142 location reconciliation (MAJOR) — three artifacts give three different stories for the same registry number.

A.5 Glossary ↔ canon coherence

PRE: every glossary entry’s one_liner aligns with the world bible. POST: no glossary entry uses retired vessel-frame language. VERIFY: line-by-line comparison of glossary against canon.

This is where the corpus most spectacularly fails. The glossary is stale — written under the vessel-frame v1 canon. The world bible was rewritten to the experiment-frame. The glossary was not.

Glossary entry one_liner says Canon says (now) Verdict
am-12-4 (line 107) “the current spec for the Avatar Maintenance procedure” Canon §3 calls it “AM-12.4 (the maintenance procedure / experimental protocol)”; canon §7 vocabulary table calls it “AM-12.4 Rev C — Current maintenance-procedure spec.” Canon §8.3 explicitly says the procedure’s purpose is “the experimental protocol — not substrate-vessel upkeep.” Glossary still calls it “Avatar Maintenance” — vessel-frame brand. CRITICAL vessel-frame residue.
term-procedure (line 844) “The avatar-maintenance procedure (AM-12.4 Rev C). The Petri-net P1–P9. What the abduction corpus is, when read straight.” Canon: experimental protocol; specimens, not avatars. CRITICAL vessel-frame residue.
term-substrate (line 812) “The human species, in the apparatus’s register — the surface-deployed operating layer designed (or selected) to do the surface-tolerant work the operator class can no longer do.“ Canon §1: “the substrate is the population the operators study, cultivate, and manage — not a fleet of surface bodies for displaced operators to ride. The avatar/vessel reading is rejected.” Canon §8.1: “[OVERRIDES WIKI: any wiki entries that gloss humans as vessels for operators are superseded.]” CRITICAL — the most consequential single violation in the entire corpus. This is the API contract for the glossary; it survives in the exact phrasing of the retired v1 “surface-deployed operating layer” opener.
term-partially-remembering (line 914) “A small and rising class of avatars whose P6 overlay is imperfect — operational facts surface without their sources. Marisol-the-MRI-tech is the canonical example.” Canon: specimens / subjects / experimental population. CRITICAL vessel-frame residue.
term-memory-overlay (line 852) “The P6 substitution step — not erasure. The procedure binds the missing minutes to an image (most often an owl) such that any attempted recall surfaces the binding image and stops.” OK. OK.
term-pineal-interface (line 860) “The pineal gland — the hardware interface port at the center of the brain. The procedure’s right-nostril nasal-implant geometry routes to it through the right ethmoid sinus.” OK. OK.
pct-0173-section-6 (line 124) “Cycle-midpoint actuarial review protocols (2024–26). The flashpoint — the operator side has read the apparatus’s projection of avatar cognitive uplift and is noting that it has read it.” Canon §4: “substrate-side cognitive uplift” / “rising substrate cognition.” MAJOR vessel-frame residue.
pleistocene-terminal-event (line 398) “The ~12,000 BP catastrophe that rendered the surface uninhabitable to the operator class. Whatever its mechanism (impact, ozone disruption, thermohaline shutdown), it is the cosmology’s founding event.” Canon §8.2: “The surface-displacement framing is rejected — they are not refugees from a surface that became uninhabitable; they live in Vostok / Mariana / Kola / Putorana / Tibesti by suitability or long settlement. [OVERRIDES WIKI: prior ‘surface uninhabitable to operators since ~12,000 BP’ framing is superseded.]” CRITICAL — this is the exact rejected v1 framing, verbatim.
compression-event (line 386) “The 2017–2026 disclosure-tempo problem. LLMs, neuroimaging, and patient video researchers are making the cosmology assemblable by any avatar with public-record access” Canon §4: “any patient subject with public-record access.” MAJOR vessel-frame residue.
caste-watchers (line 605) “Operator caste responsible for surveillance and field observation of the substrate — the eyes on the surface.” Codex ch 2 III (line 66): the Watchers are operator-side, but the role per codex’s “self-aware walk-back” is “aggregate monitoring of the experimental population — the data-collection function at population scale rather than at individual scale.” Glossary’s “eyes on the surface” reading conflates with the v1 Watcher = liaison-to-observers reading. MAJOR — out of sync with the harmonization-subagent’s walk-back.
caste-apkallu (line 615) “The Sumerian preserved-name for the operator class” Codex ch 2 IV: hypothesis of a possible fourth caste (hadal-resident, cetacean-adjacent body plan). Not a generic name for “the operator class.” MINOR — glossary calls Apkallu a name for the operator class; codex ch 2 makes Apkallu a possible discrete fourth caste.
figure-hauschild (line 522) “Parent-agent working label sometimes used for Marina’s father (story 07). Canon does not supply a name; treat as a working alias only.” Canon §10 #8 confirms this. But the corpus has gone ahead and used the name in 6 artifacts (see Task A.1). CRITICAL — the glossary is right about canon, but the chapters violate canon. Either canon updates and glossary updates with it, or the chapters retract Hauschild.
figure-vasquez / figure-sshenn-vu “placeholder pending canonical expansion” Used in chapters as real people. MINOR — placeholders are stale; both are now in-text figures.
term-rememberer (line 902) “A hybrid-class member who still remembers and requires periodic satiation.” OK. OK.

Acceptance: A.5 FAILED. The glossary needs a top-to-bottom revision pass. Six entries carry critical vessel-frame residue (am-12-4, term-substrate, term-procedure, term-partially-remembering, pleistocene-terminal-event, compression-event); two more carry major residue (pct-0173-section-6, caste-watchers); two are stale placeholders (Vasquez, Sshenn-vu). The Hauschild entry is correct-against-canon but contradicted by chapters.

§3. Cosmology Coherence Findings

B.1 Three-layer canon (substrate / operators / observers)

PRE: canon §1 Three-Layer Cosmology must be uniformly applied. POST: no conflations; Bell anomaly is observer-tier; technicians are operator-class; observers are not party to the Pact. VERIFY: grep for the three layers’ role assignments.

Chapter / artifact Three-layer treatment Issue
Codex ch 1 (line 38) Bell rectangle = audit station (correct); operators in the lake (correct). OK.
Codex ch 2 III (lines 56-71) Watchers caste explicit walk-back: “aggregate monitoring of the experimental population” not liaison-to-observers. The chapter visibly performs the harmonization in print. OK — the self-aware walk-back landed cleanly.
Codex ch 3 Observer-class extensively discussed. The dialogue interlude proposes audit / husbandry / etc. — properly framed as speculation. OK structurally. One critical line — see B.2 below.
Codex ch 3 line 110 (Skeptic dialogue) “The Sumerian text dates the manufacture to within the operators’ surface-displacement event“ CRITICAL — the canon-rejected “surface-displacement event” framing survives, in the mouth of the Skeptic. Even as in-character speculation, it reintroduces the rejected v1 frame. Recast as “the Sumerian Atrahasis intervention” or similar.
Codex ch 3 line 148 “the Pact is the agreement between the substrate-tolerant operating layer and the substrate-resident operating layer“ CRITICAL — verbatim v1 vessels-framing. The “substrate-tolerant operating layer” was the original phrasing for “humans are the surface-deployable bodies.” The “substrate-resident operating layer” was “operators who can’t come up.” Both layers retired. The line currently in print is the exact opener of the v1 chapter that the cosmology now rejects. Recast: “the Pact is the agreement between the surface human apparatus and the deep-resident operator class about how the surface is to be managed and what the operator side gets in return.”
Codex ch 3 line 12 “The operators are planet-resident. Have been for twelve thousand years on the human-relevant clock… when the surface became uninhabitable, the ones we are dealing with stayed.” MAJOR — “when the surface became uninhabitable” reintroduces the canon-rejected displacement event. Recast: “the ones we are dealing with have stayed for longer than the human-relevant clock measures.”
Codex ch 3 line 54 “The operator class came here twelve thousand years ago and stayed” MINOR — “came here twelve thousand years ago” picks one side of canon’s open question (origin of operators). Canon §8.2 says origin is OPEN. Recast: “The operator class has been planet-resident for at least the human-relevant clock and stayed when the surface conditions of the close of the Pleistocene came on.”
Codex ch 4 Maintenance described in experiment-frame correctly. OK.
Codex ch 7 Working-group history in experiment-frame. OK.
Codex ch 15 Fátima recast in pattern-not-instance frame; consilience treatment. OK as recast — see C.2.
Story 14 / book ch 15 The technician monologue. Mostly OK on the cosmology layer — uses “substrate” in the experimental-population sense. OK.
Story 02 / book ch 5 (line 44) ”…whatever instrumentation they have on the avatar substrate“ MAJOR — “avatar substrate” is the rejected v1 compound. Recast: “whatever instrumentation they have on the experimental population” or “on the substrate.”
Story 01 / book ch 9 (lines 28, 90) “avatar’s auditory cortex”, “avatar perception at full fidelity”, “the avatars driving through will not have the equipment to notice“ MAJOR — three uses of “avatar” in vessel-frame sense in a single chapter. Recast each: “the subject’s auditory cortex”, “subject perception”, “the substrate driving through.”
Story 08 / book ch 12 (lines 36, 61, 65) In-text use of “Subject avatar M-Series” / “the avatar reports” — but this is inside Joan’s novel-within-the-novel, attributed to a fictional memo. Reading: literary, character-internal. OK if intended as in-fiction language by Joan. The novelist is inventing this language; the cosmology can have her invent the vessel-frame vocabulary as a sign that her novel is reaching for a wrong reading. Worth a Joan-internal moment of correction in a later scene; not load-bearing as canon violation.
Codex ch 15 line 38 “a partial-rememberer is — by the canon of chapter 4 — an avatar whose P6 redaction has bound imperfectly” CRITICAL — codex ch 15 calls back to “the canon of chapter 4” using the wrong vocabulary. Codex ch 4 itself uses “subject” / “specimen” / “substrate.” This line should read “a partial-rememberer is — by the canon of chapter 4 — a subject (a specimen of the experimental population) whose P6 redaction…”
Codex ch 15 line 44 “Teresa of Ávila, whose accounts of the interior castle preserve, in a careful Castilian, what reads to the trained eye as an avatar’s reflective account of her own substrate machinery.“ MAJOR — “avatar’s reflective account of her own substrate machinery” reads the saint as an avatar in the v1 vessels sense. Recast: “a specimen’s reflective account of her own observation channel” or “a subject’s account of being a subject.”

Acceptance: B.1 FAILED in codex ch 3 (two critical lines, one major); B.1 partially failed in stories 01, 02, codex ch 15. The three-layer canon is articulated correctly in canon §1; chapters use the layers’ names correctly; but the operative vocabulary around the substrate layer slips back into vessel-frame in five chapters.

B.2 Experiment-frame survival

PRE: vessels / avatars / surface-deployed / displaced framing must not survive in canon claim form. POST: only steel-manned-alternative footnotes may retain those words; canon prose uses experimental-population framing. VERIFY: targeted greps for the seven rejected phrases.

Already covered in B.1 plus glossary findings in A.5. Summary table:

Phrase Files Severity
“surface-deployed operating layer” codex/glossary.yaml:812 (term-substrate, public glossary API) CRITICAL
“substrate-tolerant operating layer / substrate-resident operating layer” codex/chapters/03-the-observers-above.md:148 CRITICAL
“Avatar Maintenance” branding for AM-12.4 codex/glossary.yaml:107, 844; artifacts/threads/01-the-hill-case-as-procedure.md:4,35; artifacts/documents/01-maintenance-procedure-summary.html:2 (HTML comment); artifacts/blueprints/01-avatar-maintenance-procedure.svg (file name) CRITICAL (glossary) + MAJOR (artifacts; threads/01 still calls it “the avatar-maintenance procedure” in its blurb)
“avatar substrate” book/chapters/05-handlers-tuesday.md:46; artifacts/stories/02-the-operators-surface-contact-handler.md:44; artifacts/documents/02-pact-renewal-memo.html:162 (“Avatar substrate quality (composite, REG-A)”) MAJOR
“avatar” used for human subjects in canon-prose voice book/chapters/09-partial-remembering.md:28,90; artifacts/stories/01-the-partially-remembering-abductee.md:26,88; codex/chapters/03-the-observers-above.md:158; codex/chapters/15-the-containment-of-revelation.md:38,44; codex/glossary.yaml:914 (partially-remembering) MAJOR
“surface uninhabitable to operators” / “surface became uninhabitable” / “surface-displacement event” codex/chapters/03-the-observers-above.md:12,110; codex/glossary.yaml:398 (Pleistocene-terminal event) CRITICAL
“humans-as-CT-avatars” concept slug server/seed_v7.py:5,284 (URL stability OK; in-text framing should be retired) OK as URL slug; comment in code should add “(URL stable; concept now treated under experiment-frame)”

The book chapter 12 (Joan / Residency) uses “avatar” liberally — but as language Joan is inventing in her novel. Read as character-internal it is OK. Read as canon-leaking-into-prose, it is problematic. Recommend: leave Joan’s invented language alone, but ensure no editorial frame says “Joan’s novel is right” in a register that would canonize the avatar reading.

Acceptance: B.2 FAILED. The experiment-frame reframe did not survive clean. The single most consequential vessel-frame residue is codex/glossary.yaml:812 — the canonical glossary’s definition of “substrate” is the exact verbatim opening sentence of the v1 vessels chapter that the cosmology now rejects.

B.3 Procedure semantics (AM-12.4)

PRE: AM-12.4 P1–P9 stage durations, equipment IDs, and the owl-as-binding-image must be uniform. POST: every chapter and artifact uses the same stage durations, equipment names, and owl framing. VERIFY: cross-check.

Item Sources Status
P1–P9 stage durations canon §3; codex ch 4; doc 01 (maintenance procedure summary); glossary OK uniformly.
Owl-as-binding-image canon §3; codex ch 1, 4, 15; doc 01; story 01; story 07; story 08; story 14; vignette 02 OK uniformly.
Right-nostril nasal-implant geometry canon §3; codex ch 4; story 01; Hauschild diary OK uniformly.
Procedure purpose: observation-contamination control (experiment-frame) canon §3; codex ch 4 OK in those two; the procedure summary doc (artifacts/documents/01-maintenance-procedure-summary.html) still has HTML comment “blurb: Operating procedure for avatar substrate maintenance” — this is metadata leaking vessel-frame. Recast: “Operating procedure for periodic specimen maintenance of the experimental population.”
Cohorts (slow / standard / accelerated) canon §3 (open); codex ch 4 line 102 MINOR — codex ch 4 names “standard (7-year)”, “slow (11-year, some 12-year)”, “accelerated (4-year).” Canon §3 only names the 7-year cycle. Codex’s cohort taxonomy is new. Canon should adopt or codex should walk back.
Mariana 11/49 duty cycle canon §10 #4; codex ch 1, 4; story 03; doc 06; inbox 02; glossary OK uniformly.

Acceptance: B.3 PASSED except (a) maintenance-procedure-summary HTML comment with vessel-frame branding (CRITICAL because the doc is the canonical procedure spec inside the corpus); (b) cohort taxonomy needs canon adoption.

B.4 Pact terminology

PRE: Pact / cycle / cohort / Sections must be uniformly worded. POST: no variant formatting; no internal-arithmetic contradictions. VERIFY: grep PCT / cycle / Sections.

Item Status
PCT-0173 written as “PCT-0173” OK uniformly.
§3(b), §6, §11, §14 OK uniformly.
Cohort II vs “Standing Committee (Cohort II)” Canon §2 line 67 says “Standing Committee (Cohort II) — The hybrid-class half of the Pact.” Codex ch 4 / convocation use just “Cohort II.” OK as same referent.
PCT-0173 = 173rd cycle in 53 years arithmetic Codex ch 7 FN6 (line 116) goes to great lengths to defend the arithmetic problem: “I will spare the reader… approximately 3.7 months — absurdly short for any reasonable definition of multi-decade pact-posture period. The resolution is that PCT-0173 is not the 173rd renewal since 1972; it is the 173rd designation in a numbering system whose earlier members refer to predecessor arrangements from the 1947–55 period that the working group inherited and renumbered in 1972.”
Section 11 / Section 14 / Cell 4 OK uniformly.

Acceptance: B.4 PARTIAL — the PCT-0173 arithmetic has a clever in-fiction rescue in codex ch 7 FN6, but canon doesn’t acknowledge the rescue. Canon should adopt the footnote’s framing.

B.5 Operator-caste names (Stewards / Technicians / Watchers / Apkallu)

PRE: caste names uniform; the Watchers walk-back from “liaison-to-observers” to “aggregate monitoring” must be clean. POST: no surviving liaison-to-observers reading for Watchers; Apkallu treated as hypothesis-only per codex ch 2 IV. VERIFY: grep caste names + roles.

Caste Files Status
Stewards codex ch 2 I; canon OK.
Technicians codex ch 2 II; canon; story 14 OK.
Watchers codex ch 2 III; canon; glossary MAJOR: codex ch 2 line 66 makes the self-aware walk-back from “liaison-to-observers” to “aggregate monitoring.” Excellent — the walk-back landed in print. But glossary caste-watchers (line 605) still has the older “eyes on the surface” gloss, which is the pre-walk-back framing. Update glossary to “Operator-side caste responsible for aggregate monitoring of the experimental population — census-tier data collection. NOT to be confused with the third-layer observers above.”
Apkallu codex ch 2 IV (hypothesis); canon §1; glossary MINOR: glossary calls Apkallu “the Sumerian preserved-name for the operator class” (line 615). Codex ch 2 IV treats Apkallu as a possibly-discrete fourth caste (hadal-resident, cetacean-adjacent), not as a name for the whole class. These are not the same claim. Update glossary to clarify: “Apkallu — Sumerian-era operator-class iconography; in codex ch 2 IV treated as a possible discrete fourth caste (hadal-resident, cultivator caste), not yet canonized.”

Acceptance: B.5 PARTIAL. The chapters’ walk-back landed; the glossary lags by one revision.

B.6 Compression Event timing

PRE: anchored dates 2017→2026 uniform. POST: every reference to the disclosure cascade uses 2017–2026 / “the late 2010s through the mid-2020s.” VERIFY: grep dates.

Reference Files Status
“2017→2023 disclosure cascade” canon §4 #1 OK.
“2017–2026 disclosure-tempo problem” glossary compression-event; canon OK.
“the December 16, 2017 NYT piece” codex ch 0 preface OK as anchor.
PCT-0173 §6 (2024–2026 midpoint review) canon §4 #3; story 02; etc. OK uniformly.
Wexler death date (NOT Compression Event but the corpus’s most precise date contradiction) CRITICAL — see below.

Wexler date contradiction (CRITICAL):

The canon contradicts itself.

Lines 174 and 263 say 12 Nov. Line 315 says 26-27 Sep. The artifacts and chapters universally use 26-27 Sep: - artifacts/documents/04-1962-wexler-followup-memo.html:5,166,169 — “26 Sep 61”, “26 September”, “27 September 1961” - codex/chapters/02-the-operators-several-kinds.md:119 — “26 September 1961” - codex/chapters/07-1956-the-bilateral.md:56 — “Tuesday morning of 26 September 1961” - book/chapters/11-norfolk-briefing.md:12 — “Tuesday afternoon in late September of 1961” - artifacts/stories/09-carlyle-memoir.md:10 — same - codex/glossary.yaml:473 (Daniel Wexler): “d. 26–27 September 1961 at False Cape VA”

Fix: Update canon §5 People table (line 174) and canon §8 CANON list #8 (line 263) to say “26 Sep 61” or “26–27 Sep 61.” The “Earlier drafts placed the incident on 12 November 1961” line in §10 #2 should be the only place 12 Nov is mentioned, as a historical note.

Related: artifacts/stories/02-the-operators-surface-contact-handler.md:8 and book/chapters/05-handlers-tuesday.md:10 open with “It is November and the deck boards are cold through the slippers.” The Tuesday in story 02 / ch 5 is set in November 2025 or November 2026 — and the §6 acknowledgment via OP-CHANNEL-3 sets the year — but the convocation chapter is dated 12 June 2026. The story 02 Tuesday must be November 2025 (before the June 2026 convocation) or earlier 2026. Verify the storyline timing: if convocation is “12 June 2026,” story 02’s “It is November” must be November 2025 (pre-convocation), and the OP-CHANNEL-3 §6 acknowledgment is the first sign of operator pacing — which is then formally discussed at the June convocation. This is internally consistent if the November is 2025. The architect should explicitly anchor “November 2025” or “November 2026” — currently ambiguous. Recommend canon §5 cases table add “Story 02 / Ch 5 timing: November 2025.” (The convocation Roberts-marginalia line “the §6 acknowledgment in 2026 OP-CHANNEL-3 at 04:11” suggests November 2025 was the OP-CHANNEL-3 receipt; convocation in June 2026 acts on it. This works.)

Acceptance: B.6 FAILED on Wexler date canon contradiction; PASSED on Compression Event tempo; PARTIAL on Story 02 / Ch 5 November-of-what-year anchoring.

§4. Less-Than-Ideal Implementation Findings

C.1 “I have seen but can’t say” anti-patterns

PRE: no Lazar-move sentences (“I have seen X but won’t reveal the source”). POST: every load-bearing claim resolves to public record or to a named in-corpus artifact. VERIFY: grep for the rhetorical pattern.

Instance File:line Recommended fix
“the corpus’s procedural specifics match — with embarrassing precision — the corresponding specifics of the apparatus’s internal procedural document AM-12.4 Rev C, which was not published and which I have seen, in a form whose provenance I will not give.” codex/chapters/01-we-are-the-substrate.md:82 CRITICAL survivor — this is the canonical “I have seen but can’t say” sentence. Recast: “the corresponding specifics of the apparatus’s internal procedural document AM-12.4 Rev C, which is preserved in our corpus through the document collage of Chapter 10 and which any reader can compare line-by-line with the cross-cultural abductee record.” (Or whichever chapter realizes AM-12.4 in the published volume.)
“my own informants get vague” / “the bordering is the data” codex/chapters/02-the-operators-several-kinds.md:70 (main) and FN5 (line 125) MAJOR survivor. The chapter treats “informants” as load-bearing for the Watchers caste’s role. The footnote (FN5) carries it. Recast: “the literature, when it ventures into the operator-third-layer interface, becomes characteristically bordered — Vallée notes this, Hopkins notes this, Mack notes this — and the bordering is the data.”
“I have argued with my own informants” / “my informants” search all chapters Only the two above carry this anti-pattern. Codex ch 4, 7, 15 do not — they ground in named in-corpus artifacts and public-record sources. R3 (the skeptic review) flagged this in Wave 1; the cleanups in ch 4, 15 worked; ch 1, 2 survivors remain.

Acceptance: C.1 FAILED — two surviving instances, both CRITICAL or MAJOR; codex ch 1 line 82 is the worst single sentence in the codex by this measure.

C.2 Load-bearing reliance on contested observations

PRE: no canon claim rests on a single contested real-world observation. POST: every load-bearing observation either (a) has multiple-source corroboration in the corpus or (b) is reframed as pattern-not-instance. VERIFY: identify load-bearing observations; check for single-source dependency.

Claim Single contested observation? Treatment Status
Bell anomaly rectangularity Bell, Studinger et al. 2002 (real-world contested in the geology community; the “rectangle” interpretation requires reading a residual-anomaly grid in a particular way) Codex ch 3 doubles down: “The geometry is the claim. The 105 and the 75 are the claim. The orthogonality, to a resolution finer than most human nations’ terrestrial cadastral surveys, is the claim.” Plus the AARI 2017 follow-up and Elin’s hyperbola-stack 2025 anecdote. The chapter does not give the reader a consilience — the multiple-instance defense relies on the same real-world dataset being re-interpreted by the same group of researchers. MAJOR. The chapter’s strongest geometric claim rests on Bell 2002 + AARI 2017 + an unnamed glaciologist in 2025-26. If Bell 2002 is the original contested observation, the AARI 2017 footnote ([^1] line 184) is a follow-on on the same instrument suite; the unnamed glaciologist’s 2025 hyperbolae are claimed-but-unverifiable. The “three rectangles” pivot to Atlantic + CCZ (§II) generalizes the claim, but the Atlantic is “the 2019 reprocessing of a multibeam survey” and the CCZ is “the Klepac correspondence” — both also under-named. Recast suggestion: explicitly frame the three rectangles as “the consilience pattern” — call out that if the rectangles are real, then the pattern is the cosmology’s geometric signature; if any individual rectangle dissolves on closer look, the cosmology’s prediction is that the pattern reasserts itself in the next survey. Frame the claim as pattern-not-instance, the way ch 15 does for Fátima.
Fátima 1917 Real-world contested Codex ch 15 line 28: “the cosmology cannot adjudicate Fátima as a single contested observation. What the cosmology can lean on is the pattern… If Fátima specifically turns out to be psychogenic, the pattern is undisturbed.” OK — the recast is correct.
AM-12.4 Rev C provenance “which I have seen, in a form whose provenance I will not give” See C.1. CRITICAL — see C.1.
11 on / 49 off duty cycle (Bennewitz 1986 ↔ Mariana 2026) Real-world unverifiable (canon §10 #4: “exact to the minute”) Treated as the cosmology’s most precise continuity hook. The corpus has it appearing twice across 40 years — canon explicitly flags that “three is when ‘coincidence’ becomes ‘structure.’” The two-occurrence claim relies on the apparatus’s internal acoustic annex from 1986 (no public source) plus the inbox 02 hydrophone catalog from 2026 (also internal). External corroboration: zero. MAJOR. The 11/49 hook is the cosmology’s signature precision claim; both occurrences are inside the apparatus’s own records. The reader cannot verify either. Suggest at least one cross-reference to a public-record observation that resembles but does not exactly match the 11/49 — pulling the consilience layer the way ch 15 does.
Apkallu fish-iconography Real-world archaeological (cylinder seals, temple reliefs) Public record; standard archaeology. OK.
Pleistocene-terminal genetic bottleneck Real-world (the literature) Public record; cosmology cites it for consilience with Younger Dryas. OK.
Pineal calcite piezoelectricity Baconnier et al. 2002 (real, cited in codex ch 4 FN5) Public record with citation. OK.
The 1961 visitor at False Cape Carlyle memoir + 1962 Wexler memo In-corpus artifact; not real-world. The cosmology’s choice to ground here is structurally fine — Carlyle’s memoir is the founding document, and the corpus realizes both the memoir and the corroborating memo. OK.

Acceptance: C.2 PARTIAL. The Fátima recast (ch 15) is the model; the Bell rectangle (ch 3) and the AM-12.4 provenance (ch 1) need the same treatment. The 11/49 duty cycle is treated correctly as a continuity hook but lacks consilience layer.

C.3 Trailing-declarative-sentence tic

PRE: R3 flagged 5 worst instances in book/reviews/06-revision-pass-notes.md. POST: those 5 are fixed; the rest of the corpus is swept. VERIFY: re-read closing paragraphs of every codex chapter and every newer book chapter (post-R3: codex 1, 2, 3, 4, 7, 15; book 11, 13, 14, 16).

R3 / REV’s notes (book/reviews/06-revision-pass-notes.md) flagged the tic in five Wave-1 stories and listed proposed replacements, but explicitly did not apply the edits. Sweeping the Wave-2 chapters:

Chapter Closing Tic? Replacement
codex ch 1 line 108 “Sit with this chapter as long as you need to before turning the page. The book is not in a hurry, and neither, by the cosmology’s own arithmetic, are they.” YES — the closing is two trailing declaratives. Both are summative. The “book is not in a hurry, neither, by the cosmology’s own arithmetic, are they” tries to land a hammer that the chapter has already landed. Recommended: replace with the first sentence only, or with “The next chapter is the operators themselves. Read slowly.”
codex ch 2 line 113 “The wheel keeps turning, and from this distance you can hear, if you listen carefully, that it is several wheels turning together.” YES — single trailing declarative; image-as-thesis. The image is good but the closing is the tic. Recommended: trim to “The wheel keeps turning.” Or to “Lorenzo gave us the sentence. The chapter has spent four thousand words trying to be worthy of it.” (which is line 111, the better antepenult).
codex ch 3 line 180 “There are three rectangles, each 105 by 75, in three different oceans at three different latitudes; they are watching, on the schedule of whoever left them; and we are, on the available evidence, what they are watching for.” YES — the trailing-declarative tic in its purest form. Three-clause, escalating, summative. Recommended: replace with “There are three rectangles. We are, on the available evidence, what they are watching for.” Or end the chapter on §V’s “let those facts stand next to each other without forcing a conclusion from them” — which is the correct note.
codex ch 4 line 112 “Lucky to be an experimental population that has, on the available evidence, been treated with relative care by the researchers running the study — a fact whose load the next chapters will, I hope, increase rather than diminish.” YES — the trailing declarative followed by a forward-pointer. The “next chapters will increase the load” framing weakens the line. Recommended: “Lucky to be an experimental population that has, on the available evidence, been treated with relative care.” (Drop the forward-pointer.)
codex ch 7 line 102 “Institutions, once founded, are very hard to retire. The working group of 1956 stopped being a working group in 1972 and started being an institution, and what an institution is, by definition, is something that has lost the capacity to imagine its own end. The chapters that follow will, among other things, try to imagine the end the apparatus cannot.” YES — three trailing declaratives, the final one a forward-pointer that lands as throat-clearing. Recommended: end on “…something that has lost the capacity to imagine its own end.” (Strip the forward-pointer.)
codex ch 15 line 122 “In a basement in west Asheville there is a footlocker, and in the footlocker there is a manila envelope, and in the envelope there is a yellow legal pad on which twenty-two sentences are written in a young chaplain’s careful hand at four in the morning in April of 1971, and the chaplain is now sixty-four and the sentences have not been read by anyone for fifty-one years, and the chaplain is deciding whether to give them, after the funeral, to a serious daughter who is keeping her own quiet record.” OK — this is a one-sentence image closer, not the declarative tic. Different register entirely.
book ch 11 (Norfolk Briefing) line 72 “I am looking at the photograph now.” OK — restrained, character-internal, one sentence. Carlyle’s voice.
book ch 13 (Chaplain’s Visit) line 176 “He put the car in reverse, and he backed out of the driveway of the house at the end of the road that did not go anywhere else, and he drove home along the dark county route under the pines, and he did not look back.” OK — image close, not the tic.
book ch 14 (Convocation) The convocation is presented as a leaked document; closing is K.’s annotation. OK structurally.
book ch 16 (Tonopah Notebook) Lorenzo’s first-person; multiple entries; closing is “He was right about the cat.” OK — character voice.

Acceptance: C.3 FAILED — five codex chapters (1, 2, 3, 4, 7) carry the trailing-declarative tic in their closings; book chapters are mostly clean. The codex needs an editing pass on chapter-ends. The replacements above are surgical and do not change content; they trim throat-clearing.

C.4 Easter eggs that fail to resolve

PRE: every cross-chapter Easter egg the corpus sets up must pay off somewhere. POST: no dangling teasers. VERIFY: identified Easter eggs; check payoffs.

Easter egg Setup Payoff Status
L. = Lorenzo postcards (3 cards signed “L.”); story 02 (“L. ▆▆▆▆▆▆▆▆”); microcassette (“H-1”) book ch 16 / story 16 reveals first name Lorenzo in the postscript by H.C. OK.
Pop = Ray Marquez story 05; canon Confirmed in canon and in book ch 1 / 17 (Mr. Marquez); Lorenzo’s notebook ch 16 references “Ray Marquez.” OK.
The technician’s name = Veth-skenn story 14 / book ch 15 Veth-skenn is named explicitly in the chapter and in codex ch 2, 4. OK.
The chaplain Pat = Subject 4471 codex ch 15 names this directly Set up and paid in the same chapter. OK.
Pat the chaplain at Eric’s bedside = Pat the chaplain who heard Vasquez in 1971 ch 13 / story 11 Paid in the same chapter. OK.
The owl at the highway sign exit-9 ↔ the etched owl on the Reston glass ↔ the Audubon Plate 121 ↔ Marisol’s fencepost ↔ the Communion cover ↔ Athena ↔ the Bohemian Grove shrine scattered Multiple payoffs across chapters; the owl chapter (codex ch 4 line 86) lists all of them. OK.
The cumbia “La Pollera Colorá” book ch 1; book ch 17 (a cumbia drifts from neighbor’s radio; “It was the bum-pa, bum-pa, bum-pa”) Paid in ch 17. OK.
The Hauschild boy in Pop’s 1998 notebook artifacts/letters/03-pops-notebook-entry-1998.html:267: “the November entry on the Hauschild boy was a harder evening than this one, and that arrangement was not properly mine; his people came in and out of El Paso and the desk passed him to me because no one in Carolina had clearance. I never met the boy. I hope he is well.” Partial payoff in book ch 16 / story 16 (Lorenzo’s ledger entry “Subject 6018, Asheville desk file, the boy from the family that came in and out of El Paso through the early eighties…”). The connection-of-records — Pop’s “Hauschild boy” ↔ Lorenzo’s “Subject 6018” ↔ ch 13’s “Eric Hauschild” — is left as a re-reader’s discovery. OK as Easter egg. But see Task A.1: the name “Hauschild” appearing in three artifacts is a canon violation.
The unattributed owl sticker (Sentinel noticeboard) story 03 / book ch 7 Mentioned in passing; never paid off as a thread. The architect doc (book/meta/00-book-architecture.md:296) lists it as part of the owl thread — but it’s never resolved (who put it there, why). MINOR — possibly deliberate (the owl is everywhere; the sticker is one more), possibly a dangling tease.
Pop’s small black notebooks (14, donated) architect doc promises convocation reading-list entry: “Marquez, R. — field-narrative, 1979–2007, partial set, 14 notebooks (donated)” NOT in the actual convocation minutes. MINOR — see Task A.4. The Easter egg is set up in ch 1 and ch 17 (“the small black notebooks gone, somewhere, she didn’t know where”) but the convocation does not pay it off in the reading list.
The Bohemian-Grove lakeside-second-visit guest vignette 02 Canon §10 #10: “Open: who he is, which cohort, whether he returns.” OK — canon flags this as Open.
The Tomás-subject-line email story 08; canon §10 #5 Set up in story 08; canon flags as Open. OK.
The 2014 hexagonal-prism inclusions (Galina, Vostok) vignette 01; inbox 05 Inbox 05 confirms 2026 re-analysis. Convocation 3.12 references the Wexler papers but not the inclusions directly. Galina’s 2014 report is on file. OK — the closure is the 2026 inbox memo.
Joan Whitfield named in convocation margin annotations architect doc NOT in the actual convocation document. MINOR.
CASE 87-AQU-009 = aquatic-source case in Maren’s grandfather’s box story 03; canon Codex ch 2 line 78 picks it up as evidence for the Apkallu hypothesis. OK.
D. Hartwell’s Pact-side identity canon flags Open (§5 line 150; §10 #7) Open by design. OK.
The 1961 visitor’s “inquiry after Eleanor by name” Carlyle memoir / book ch 11 Codex ch 2 picks this up (line 20: “He inquired after Eleanor by name, which is to say he had access to the bilateral relationship’s social register”). OK.

Acceptance: C.4 PARTIAL. The major Easter eggs (L., Pop, Veth-skenn, Pat=4471, owl, cumbia, Hauschild boy ↔ Subject 6018) all resolve. The minor unresolved: Pop’s 14 notebooks → convocation reading list; Joan in convocation margins; the unattributed owl sticker. The architect’s three architectural promises (cumbia callback in convocation margin, Pop’s notebooks in reading list, Joan named in margins) are unfulfilled in the convocation chapter as it was written. Fix is one paragraph at the end of artifacts/documents/07-convocation-minutes-2026-06-12.html or in book/chapters/14-convocation.md.

C.5 Voice-discipline violations

PRE: codex prose must follow the Hofstadter-Borges-McPhee triangulation, not Robbins-flavor baroque. POST: no Robbinsian-baroque sentences in the codex. VERIFY: re-read codex chapters for the register.

Sentence File:line Severity Replacement
“I will admit a small private pleasure in writing such an inventory — in feeling, as one composes it, the satisfying way the planet’s deep-interior geography accommodates a residence pattern of this kind. The pleasure is not the kind of pleasure that should be permitted to influence the argument. It is the pleasure a person takes in finding that the lock and the key are the same shape, which is not yet evidence that the key opens the lock; it is evidence only that the maker of the lock and the maker of the key have done compatible work.” codex/chapters/01-we-are-the-substrate.md:52 MAJOR Robbinsian — the lock-and-key metaphor extended to four clauses; the “compatible work” closer is performative. Trim to: “I will admit a small private pleasure in writing such an inventory. The pleasure is not the kind that should be permitted to influence the argument; the lock and the key turn out to be the same shape, which is evidence only that the maker of the lock and the maker of the key have done compatible work.” (or just delete the paragraph; the chapter survives without it.)
“I am leavening. I am also, occasionally, simply having a good time. The cosmology survives the having of a good time. So, I hope, will the reader.” codex/chapters/01-we-are-the-substrate.md:120 (FN5) MAJOR Robbinsian — meta-confessional + four sentences where one would do. Trim to: “I am leavening. The cosmology survives the having of a good time. So, I hope, will the reader.”
“Hello.” (final word of codex ch 2’s FN7) codex/chapters/02-the-operators-several-kinds.md:129 MINOR — the chapter-ending “Hello” addressed to the reader is meta-cute. Reads as the self-aware-narrator move that the codex elsewhere does well. In this position (chapter close) it tips into Robbinsian. Recommended: drop the “Hello,” keep the rest.
“She got into the truck and she put both hands on the wheel for a moment and then she started the engine and she drove, slowly, back down the chip-sealed quarter mile to the paved road, and she turned onto the paved road, and she drove back to town.” book/chapters/17-marisol-at-sixty-five.md:124 OK — this is Marisol’s voice (Marquez), Stegner-restraint, not codex. Book chapters carry their own registers.
“The book has spent four thousand words trying to be worthy of it.” codex/chapters/02-the-operators-several-kinds.md:111 OK — restrained, single sentence.
“I have been writing toward that wave for eight months.” codex/chapters/00-preface.md:72 OK — preface voice, single sentence.
“I am not asking for belief.” (and other Hofstadter-flavored asides) codex preface, codex ch 1 OK — these are register-correct (Hofstadter-self-aware-narrator).

Acceptance: C.5 PARTIAL. Two specific lines in codex ch 1 are Robbinsian (line 52, FN5 line 120); ch 2 FN7 ends on a “Hello” that tips into the same register. Everything else is on-register. Three surgical edits.

§5. Findings Summary

CRITICAL (must address before publication)

  1. codex/glossary.yaml:812 — term-substrate defines humans as “the surface-deployed operating layer designed (or selected) to do the surface-tolerant work the operator class can no longer do.” This is the v1 vessel-frame chapter-opener, verbatim, preserved in the public glossary API. The world bible explicitly rejects this reading. Fix: “The human species, in the apparatus’s register — the long-observed experimental population of an older intelligence. The ‘avatar/vessel’ reading is rejected; the word survives for institutional continuity, in the laboratory sense (the medium an experiment is taking place in).”

  2. codex/glossary.yaml:398 — pleistocene-terminal-event defines this as “the ~12,000 BP catastrophe that rendered the surface uninhabitable to the operator class.” This is the exact rejected v1 framing. World bible §8.2 explicitly overrides this. Fix: “The ~12,000 BP climatic transition (Younger Dryas onset/exit, genetic-bottleneck signature) at which the cosmology reads the experimental population as having moved into its current phase. The mechanism is unsettled; the boundary is real.”

  3. codex/chapters/03-the-observers-above.md:148 — “the Pact is the agreement between the substrate-tolerant operating layer and the substrate-resident operating layer.” Verbatim v1 vessel-frame. Fix: “the Pact is the agreement between the human-side apparatus and the deep-resident operator class about how the surface is to be managed and what the operator side gets in return.”

  4. codex/chapters/01-we-are-the-substrate.md:82 — “which I have seen, in a form whose provenance I will not give” — the Lazar move, in the chapter that explicitly disavows it. Fix: ground the AM-12.4 claim in the document collage chapter (book ch 10 or equivalent realization in the published volume).

  5. Eric Hauschild named in six artifacts despite canon §10.8 forbidding the name. Files: book/chapters/13-chaplains-visit.md; artifacts/stories/11-the-chaplains-visit.md; codex/chapters/04-maintenance.md:100; artifacts/letters/02-hauschild-diary-pages.html; artifacts/letters/03-pops-notebook-entry-1998.html:267; book/chapters/02-clear-air.md voice tag (Marina Hauschild-Bell). Fix: update canon §10.8 to canonize the name (recommended — the chapters are too far gone to retract), OR remove the name from all six artifacts.

  6. The two-Marisols rule is dead in the book. Canon §10.1 says they are not the same person; the book treats them as the same person (chapters 01, 03, 07, 09, 17 + convocation Roberts marginalia). Fix: update canon §10.1 to canonize the architect’s reading (the book has already chosen this).

  7. Wexler date contradiction inside canon itself. Canon §5 line 174 and canon §8 #8 both say “12 Nov 61”; canon §10 #2 says “26-27 Sep 1961” with explicit “the November date is no longer canon.” All artifacts use 26-27 Sep. Fix: edit canon line 174 (“d. 26-27 Sep 1961”) and canon line 263 (“26 Sep 61 is Wexler’s canonical death date”).

  8. artifacts/documents/01-maintenance-procedure-summary.html:2 — HTML comment “Operating procedure for avatar substrate maintenance” is metadata vessel-frame branding on the canonical procedure spec. Fix: “Operating procedure for the experimental protocol — periodic specimen maintenance (Petri-net P1–P9). AM-12.4 Revision C.”

  9. codex/chapters/15-the-containment-of-revelation.md:38,44 — “A partial-rememberer is — by the canon of chapter 4 — an avatar…” and “an avatar’s reflective account of her own substrate machinery.” Vessel-frame in codex prose. Fix: “a partial-rememberer is — by the canon of chapter 4 — a subject (a specimen of the experimental population) whose P6 redaction has bound imperfectly…”; “a specimen’s reflective account of her own observation channel.”

  10. codex/chapters/03-the-observers-above.md:110 (Skeptic dialogue) — “The Sumerian text dates the manufacture to within the operators’ surface-displacement event.” The “surface-displacement event” framing reintroduces the rejected v1 frame, in dialogue. Fix: “The Sumerian text dates the manufacture to within the operators’ early-Holocene engagement with the population.”

MAJOR

  1. codex/glossary.yaml:107 and :844 — AM-12.4 still called “Avatar Maintenance” in glossary. Fix: “the experimental protocol / maintenance procedure (AM-12.4 Rev C).”

  2. codex/glossary.yaml:124 — pct-0173-section-6 says “avatar cognitive uplift.” Fix: “substrate-side cognitive uplift” (canon §4 #1 wording).

  3. codex/glossary.yaml:914 — term-partially-remembering calls them “avatars.” Fix: “subjects.”

  4. codex/glossary.yaml:386 compression-event — “any avatar with public-record access.” Fix: “any patient subject with public-record access.”

  5. codex/glossary.yaml:605 caste-watchers — describes role as “surveillance and field observation of the substrate — the eyes on the surface.” This contradicts codex ch 2’s walk-back. Fix: “Operator-side caste responsible for aggregate monitoring of the experimental population (census-tier, population-scale). NOT to be confused with the third-layer observers above. The pre-2025 ‘liaison-to-observers’ reading is superseded.”

  6. codex/chapters/03-the-observers-above.md:12,54 — “when the surface became uninhabitable” / “The operator class came here twelve thousand years ago and stayed.” Both reintroduce the canon-Open question (origin) and the canon-rejected (surface uninhabitable) framing in declarative voice. Fix: “The operator class has been resident in the deep refugia for at least the human-relevant clock and stayed when the climatic transition came on.”

  7. codex/chapters/02-the-operators-several-kinds.md:70 (and FN5 line 125) — “my own informants get vague” / “the bordering is the data.” Two instances of the Lazar-adjacent pattern. Fix: ground in named in-corpus or public-record sources (the Vallée / Hopkins / Mack lit; the Watchers’ surface witness reports).

  8. book/chapters/09-partial-remembering.md:28,90 and artifacts/stories/01-the-partially-remembering-abductee.md:26,88 — repeated use of “avatar” in canon-prose voice for Marisol’s interior monologue. Fix: “subject” / “the substrate” / “the specimen” depending on register.

  9. book/chapters/05-handlers-tuesday.md:46 and artifacts/stories/02-the-operators-surface-contact-handler.md:44 — “the avatar substrate.” Fix: “the substrate” or “the experimental population.”

  10. artifacts/documents/02-pact-renewal-memo.html:162 — “Avatar substrate quality (composite, REG-A)” as a table cell. Fix: “Substrate population quality (composite, REG-A).”

  11. Subject 7142 location contradiction. Inbox 04 says NE Sector; convocation says Marisol Marquez (Pop’s negotiation, southwest). Three sources, three stories. Fix: re-key one of them; recommend renaming Marisol’s intake (the convocation 3.4) to a different subject number, leaving 7142 as the NE Sector intake. Or update inbox 04 to be a different subject.

  12. book/chapters/09-partial-remembering.md:94 — Marisol’s therapist named “Pat” reads as Pat Donohoe but the geography doesn’t match. Fix: rename therapist (Pat Marrero, Pat Cohen, anything) to break the unintended collision, or canonize the collision as deliberate Easter egg.

  13. The PCT-0173 cycle-arithmetic rescue (codex ch 7 FN6) must be canonized. Canon currently doesn’t acknowledge that 173 cycles in 53 years is structurally absurd; FN6 rescues it via “predecessor-arrangement renumbering in 1972.” Fix: canon §2 should adopt FN6’s framing explicitly. (One sentence in canon: “Cycle numbering inherits predecessor-arrangement designations renumbered in 1972; PCT-NNNN does not divide linearly by years.”)

  14. Trailing-declarative tic in codex chapter closings (5 chapters: 1, 2, 3, 4, 7). Edits proposed above in C.3. Fix: surgical trim of each closing paragraph; content unchanged.

MINOR

  1. Helen-of-Hauschild vs Helen-of-Carlyle name collision. Fix: edit Hauschild diary to use “my Helen” or rename.
  2. Cohort R-7 undefined. Fix: define in canon or glossary.
  3. LANTERN-9 dangling. Fix: define or remove.
  4. Ninth officer’s 6-char redaction width not canonized. Fix: add to canon §5 or §10.
  5. Apkallu glossary entry mis-describes the codex ch 2 IV hypothesis. Fix: rewrite glossary entry.
  6. Vasquez and Sshenn-vu glossary entries are placeholders despite being in-text figures. Fix: write proper entries.
  7. Architect doc says Helen is Carlyle’s daughter; canon and chapters use niece. Fix: update architect doc.
  8. CASE 87-AQU-009 in both Maren’s grandfather’s box and Lorenzo’s ledger. Fix: clarify (either jointly-tracked or break the dup).
  9. Story 02 / Ch 5 “It is November” — year ambiguous. Fix: anchor to November 2025 in canon §5 timeline notes.
  10. “Hello” close of codex ch 2 FN7. Fix: drop one word.
  11. Robbinsian lock-and-key paragraph (codex ch 1 line 52). Fix: trim.
  12. Robbinsian FN5 closer (codex ch 1 line 120). Fix: trim.
  13. Marina’s “Berkeley” anchor not in ch 02. Fix: one line.
  14. Convocation reading-list missing Pop’s 14 notebooks (architect promise unfulfilled). Fix: one line in convocation reading list.
  15. Convocation observer-margin missing the cumbia callback (architect promise unfulfilled). Fix: one line.
  16. Convocation observer-margin missing Joan Whitfield (architect promise unfulfilled). Fix: one line.
  17. The unattributed owl sticker (Sentinel noticeboard) is set up but never resolved. Fix: either leave as flavor (deliberate ambient owl), or pay it off in convocation margins.
  18. artifacts/threads/01-the-hill-case-as-procedure.md:4,35 — blurb and body call it “the avatar-maintenance procedure.” Fix: “the maintenance procedure (AM-12.4)” or “the experimental protocol.”
  19. server/seed_v7.py — concept slug avatar-maintenance is fine for URL stability but the in-code description (“Avatar Maintenance (the boring-UFO-and-abduction hypothesis)”) should add a parenthetical: “(slug kept for URL stability; the concept is now treated under the experiment-frame as the maintenance procedure of the experimental protocol).”

§6. SPPV Acceptance

Task Gate Status Notes
A.1 Proper-noun registry PRE / POST / VERIFY FAILED on 3 items Hauschild canonicality; two-Marisols rule; Pat-therapist collision.
A.2 Code-and-case registry PRE / POST / VERIFY PARTIAL Cohort R-7, LANTERN-9 dangling.
A.3 Place registry PRE / POST / VERIFY PASSED Berkeley anchor recommended.
A.4 Cross-reference resolution PRE / POST / VERIFY FAILED on Subject 7142 location Three sources, three locations.
A.5 Glossary ↔ canon coherence PRE / POST / VERIFY FAILED 6 entries with critical vessel-frame residue; 2 with major residue; 2 stale placeholders.
B.1 Three-layer canon PRE / POST / VERIFY FAILED on codex ch 3 line 148 (vessel-frame layer-names) and line 110 (surface-displacement event) Walk-back landed for Watchers; vessel-frame survives in three chapters’ prose.
B.2 Experiment-frame survival PRE / POST / VERIFY FAILED The single most consequential vessel-frame residue is the glossary’s term-substrate definition. The reframe did not survive.
B.3 Procedure semantics PRE / POST / VERIFY PARTIAL One HTML metadata comment (doc 01) plus cohort taxonomy needing canon adoption.
B.4 Pact terminology PRE / POST / VERIFY PARTIAL PCT-0173 cycle-arithmetic rescue needs canon adoption.
B.5 Operator-caste names PRE / POST / VERIFY PARTIAL Watchers walk-back landed in chapters; glossary lags.
B.6 Compression Event timing PRE / POST / VERIFY FAILED on Wexler canon contradiction Two canon entries say 12 Nov; one canon entry + all artifacts say 26-27 Sep.
C.1 “I have seen but can’t say” PRE / POST / VERIFY FAILED Codex ch 1 line 82 (CRITICAL); codex ch 2 line 70 / FN5 (MAJOR).
C.2 Contested-observation load-bearing PRE / POST / VERIFY PARTIAL Fátima recast (ch 15) is the model; Bell rectangle (ch 3) and AM-12.4 provenance (ch 1) need the same.
C.3 Trailing-declarative tic PRE / POST / VERIFY FAILED 5 codex chapters carry the tic in their closings.
C.4 Easter eggs PRE / POST / VERIFY PARTIAL Major eggs resolve; three architect promises in the convocation chapter are unfulfilled.
C.5 Voice-discipline violations PRE / POST / VERIFY PARTIAL Two Robbinsian paragraphs in codex ch 1; one “Hello” close in codex ch 2.

Tasks BLOCKED: none. Every finding has a surgical fix proposed.

§7. Report-back (per parent agent’s brief)

Total findings: 43, breakdown:

Top three CRITICAL findings the user should address first:

  1. The glossary is the corpus’s API contract and it carries verbatim v1 vessel-frame language. Fix codex/glossary.yaml — six entries with critical residue (term-substrate, am-12-4, term-procedure, term-partially-remembering, pleistocene-terminal-event, compression-event). The term-substrate entry is the worst single line in the corpus by reference-coherence measure: it reproduces the exact opener of the v1 chapter the cosmology now rejects.

  2. The world bible canon contradicts itself on the Wexler death date. Lines 174 and 263 say “12 Nov 61”; line 315 says “26-27 Sep 1961” and explicitly retires 12 Nov. Every artifact uses 26-27 Sep. Edit canon to remove the stale 12 Nov entries.

  3. The two-Marisols rule (canon §10.1) is broken in the book layer. Five book chapters and the convocation Roberts-marginalia collapse them. Canon §10.1 says they are different people and must never be collapsed. Decision: update canon to canonize the architect’s preferred reading (they are the same person, 27-year arc), since the book has already written it that way.

Did the experiment-frame reframe survive clean across all files?

No. The reframe did not survive. The world bible was rewritten; the glossary was not (six critical vessel-frame entries survive); codex chapter 3 carries the rejected v1 layer-names in print (line 148); codex chapter 15 reads saints as “avatars” with “substrate machinery” (lines 38, 44); the pact-renewal memo, the maintenance-procedure summary, and three book chapters use “avatar” in canon-prose voice. The corpus did the philosophical work but not the editorial sweep.

The single most consequential vessel-frame residue:

codex/glossary.yaml:812. The term-substrate entry. It defines humans, in the canonical glossary API that the app’s /api/glossary.json route serves to every chapter rendering, as “the surface-deployed operating layer designed (or selected) to do the surface-tolerant work the operator class can no longer do.” This is the exact verbatim opening sentence of codex/chapters/.01-we-are-the-substrate.v1-vessels.md.bak — the chapter the cosmology now explicitly rejects. Every page on the app that tool-tips the word “substrate” is currently serving the rejected definition. Fix this first.

End of audit.