systems/migration_integrity_scan.md
Date: 2026-06-28. Tool: harness/scan_migration_integrity.py (+ refined prose sweeps).
Scope: every markdown file under _source/. Read-only; no source bytes changed.
Two genuine front-truncation corruptions exist (a section/bullet whose front was
deleted, leaving a lowercase mid-word fragment — the "ouncil records" signature):
| # | File:line | Surviving fragment | Status |
|---|---|---|---|
| 1 | RCM:5597 | * ouncil records. Status: owned digital free AU public access. | KNOWN — Ch 59 Africa Revisit; clean source supplied; patch staged (see below). |
| 2 | RCM:6472 | * tite soul reception scholarship | NOVEL — needs triage before any further work. |
RCM = _source/01_Tier_1_Foundation/T1_Research_Corpus_Master [DRAFT v1.1.4b].md.
Context: a consolidated cross-listing block at the tail of the Ch 69 Maatherion
section (lines 6456–6500) that carries Ch 70–74 endgame substrate. The truncated
bullet sits directly above its siblings:
6471 * Sacred Texts Hindu Mythology Vedic and Puranic ... cross-listed Ch 16 ... 6472 * tite soul reception scholarship <-- TRUNCATED 6473 * Plato Republic Books 4 9 tripartite soul Republic-Phaedrus comparative 6474 * Vlastos Platonic Studies Charioteer reception
tite is a non-word fragment; the neighbouring bullet names the **tripar·tite
soul**, and the whole cluster (Plato, Vlastos, Campbell threshold-guardian,
Federal Reserve/Eichengreen banking, Loki-binding, Sauron) is **Ch 74 Cassius
Velheim Charioteer substrate**. Almost certainly the front of a bullet such as
* [Plato Phaedrus 246 tripar]tite soul reception scholarship was deleted.
**Hypothesis only — the original text is not recoverable from the repo. This is a
[DECISION NEEDED - JOSH]: supply the clean bullet from the recovered source, the
same way Ch 59 was supplied.**
CVD:1188 (### between §19 intro and §19.1) and T3_Core_Characters:973 (trailing # at end of file). Both are stray empty
heading markers from the gdoc→md export; no content loss. Cosmetic only.
## Ch 3, ### Tier 2, Schema_Dictionary ## §7). A known, pervasive gdoc list-nesting artifact in the
RCM and Schema Dictionary; not corruption (this is why a plain ^## grep
under-counts RCM chapters).
ownership-status keywords (* owned physical, * needed Tier 1), real lowercase
surnames (van Bemmelen, de Silva, deMenocal), JSTOR query terms in quotes
(* "semangat" + "Malay" ...), version-log lines (v1.1.2 ...), and the shell/
Python code blocks in T99_Project_2_Book_Processing_Workflow. After excluding
these, only the two truncation fragments above remain.
### Subsection directly under ## Section with no intervening body). Not corruption.
Compared the surviving Ch 59 back-half (RCM 5597–5627) against the clean Ch 59
source supplied. They are identical past the cut point: repo 5597 `ouncil
records... is the tail of the clean Tier-3 bullet ...Peace and Security
C|ouncil records.`; repo 5598–5600 = clean Tier-3 bullets 6–8; repo 5602–5611 =
clean JSTOR (8); 5613–5621 = clean Free-source (7); 5623–5627 = clean Sacred Texts
USB (3) — all exact. No divergence beyond the single mid-word break → **no
additional drift in Ch 59.** Deleted front-half = ## Ch 59 header + Verified
canon + all Tier 1 + all Tier 2 + Tier-3 bullets 1–4 + front of bullet 5.
Patch is staged (replace RCM 5597–5627 with the clean section, normalised to
neighbour formatting: ## … / ### … headers, * bullets, and the
doc's pandoc escaping \_ / digit-trailing \.). Held — not written, pending
(a) your go-ahead on the diff and (b) resolution of the novel RCM:6472 hit.
Triage RCM:6472 first (supply the clean bullet). Then a single combined RCM repair
commit can land both fixes. Until then: do not proceed to Phase 5.5 (T2 deletion),
ch_cross appends, or Chunk 5B-2 drafting.