10 - Direct Blueprint Revisions Applied

All revisions were applied directly to the blueprint KB documents under knowledge/dev/reports/architecture/t1-fix7-existing-system-refactor-execution-blueprint-2026-06-08/ via patch_document (targeted, section-level). Production was READ-ONLY; no SQL was applied.

blueprint doc	rev	what changed
00 readme-first	3 → 6	DOC_STATUS: ACTIVE_AUTHORITY + boundary-home marker; new §Active-authority boundary registry (ACTIVE/SUPERSEDED tables); central-finding step 3 reworded to the uniform end-state; stale self-verdict replaced with the current status + fenced pre-Option-Beta history
01 inventory	5 → 6	DOC_STATUS: ACTIVE_AUTHORITY marker
02 design-to-live mapping	24 → 33	DOC_STATUS + recheck-3 note; §H rewritten to the three-set model — §H intro (typed-universe table + U_legacy ≡ U_legacy_object), §H.2 (object-only membership), §H.3 (uniform end-state over objects), §H.4 split into §H.4.A object denominator / §H.4.B object × principal privilege join / §H.4.C entry-vector coverage, §H.5 (establishment order: object membership then privilege then coverage)
03 gap classification	4 → 6	DOC_STATUS marker; recheck-3 note (three-set separation for G-11/G-12/G-13; boundary scopes the no-disposition guard)
04 construction order	37 → 39	DOC_STATUS + recheck-3 note; S19 guard count 42 → 47 incl. the 5 new guards
05 rollback	23 → 24	DOC_STATUS + recheck-3 note (snapshot/verification over object set + separate principal universe)
06 test/guard	30 → 38	DOC_STATUS + recheck-3 note; G-LEGACY-TARGET-CLOSED-DENOMINATOR re-scoped object-shape; G-LEGACY-NO-DISPOSITION-AUTHORITY re-scoped to ACTIVE_AUTHORITY; +5 guard rows (G-U-LEGACY-OBJECT-ONLY, G-PRINCIPAL-SET-SEPARATE, G-ENTRY-VECTOR-SEPARATE, G-ACTIVE-AUTHORITY-SCOPE, G-NO-SUPERSEDED-CONSUMPTION); guard-quality rules 7+8; total 42 → 47; +recheck-3 negative tests
07 package split	36 → 40	DOC_STATUS + recheck-3 note; sequencing bullet (set separation + active-authority scope); PKG-D validation adds the new guards; PKG-I count 42 → 47
08 hard blocks	16 → 17	DOC_STATUS + recheck-3 note (birth gateway/DOT/scheduler = U_entry_vector/boundary, never object members)
09 XHigh review	1 → 2	DOC_STATUS: SUPERSEDED_NON_AUTHORITY marker (historical report)
10 Max review	1 → 2	DOC_STATUS: SUPERSEDED_NON_AUTHORITY marker (historical report)
11 revisions-from-reviews	1 → 2	DOC_STATUS: SUPERSEDED_NON_AUTHORITY marker (historical report)
12 final verdict	18 → 29	DOC_STATUS + boundary note; Verdict → ..._AFTER_CODEX_RECHECK_3_READY_FOR_CODEX_RECHECK_4; new Codex recheck-3 set-separation patch pass section + "why recheck-4" reasoning; historical RECHECK-2/owner-semantics/critical-review/A-K + Independent XHigh/Max sections fenced as SUPERSEDED_NON_AUTHORITY; explicit-asks → recheck 4 (+asks 6 set-separation, 7 boundary); track-summary + blocking-status guard count 42 → 47; next = recheck 4
checkpoint (blueprint)	9 → 17	DOC_STATUS marker; final status → recheck-4-ready; recheck-3 patch-pass section; pre-recheck-3 pass narratives + Phase outcomes fenced; central finding reworded; Next → recheck 4

Edit method note

patch_document (exact single-occurrence replace) was used throughout so the large unchanged majority of each doc (e.g. doc 02 §A–§G, doc 06's full guard table) stayed byte-identical; only the changed regions were rewritten. Same-doc patches were applied sequentially (read-modify-write); cross-doc header patches were batched in parallel.