RS5B-PATCH1-03 — RS5B Document Impact Map and Superseded Wording — 2026-06-21

Macro: RS5B-PATCH1 · Deliverable: 03 of 7 · design-only. Method: I re-read all nine RS5B directory documents plus the RS5B rollup in full and scanned them for the target phrases. Classification per occurrence: OK / SUPERSEDED / NEEDS_PATCH. Rule: patch only the contradiction; do not rewrite unrelated RS5B sections.

1. Phrase scan (across the 4 in-scope RS5B docs)

phrase	RS5B-05	RS5B-07	RS5B codex packet	RS5B rollup
kept out of … authorization binding	1× (item 5) → SUPERSEDED	absent	absent	absent
effect_identity out of binding	absent	absent	absent	absent
approval not bound (APPROVAL_NOT_BOUND_TO_EFFECT_IDENTITY)	1× (item 5 reject code) → OK (correct enforcement)	absent	absent	absent
authorization intent	1× (item 5 element name) → NEEDS_PATCH (naming clarified)	absent	absent	absent
necessary-not-sufficient	1× (authority note, about packet vs Chairman) → OK (correct usage)	absent	absent	absent
effect_identity (as a listed packet element)	item 5	§1 summary row → OK	§3 (13 elements) → OK	§4 bullet → OK

Net: exactly one direct unsafe occurrence — RS5B-05 item 5's requirement phrase. Two shorthand occurrences require clarification by this addendum: RS5B-05's element name effect_identity / authorization intent and RS5B-03 Model D's required-evidence phrase effect/authorization intent. They mean two explicit elements: pure effect_identity plus authorization_binding_digest containing it. Other documents list effect_identity without asserting it is outside the binding and therefore inherit the corrected contract by reference.

2. Per-document classification

document	occurrence	classification	action
RS5B-03	Model D required evidence: "effect/authorization intent"	NEEDS_PATCH (minor)	superseded reading = pure effect_identity + authorization_binding_digest containing it
RS5B-05	item 5 requirement: "effect_identity … kept out of the authorization binding (separation)"	SUPERSEDED	replaced authoritatively by [[02-corrected-effect-identity-and-authorization-binding-contract]] §5
RS5B-05	item 5 element name "effect_identity / authorization intent"	NEEDS_PATCH (minor)	corrected element name = "effect_identity (pure) + authorization binding" (PATCH1-02 §5); reject code unchanged
RS5B-05	item 5 reject APPROVAL_NOT_BOUND_TO_EFFECT_IDENTITY	OK	kept; +AUTHORIZATION_BINDING_MISSING_EFFECT, EFFECT_IDENTITY_IMPURE added (PATCH1-02 §3)
RS5B-07	§1 row "Authorization packet — 13 conjunctive elements … none satisfied today"	OK	inherits corrected item 5; no rewrite
RS5B-07	§3 PASS-level table, §5 "necessary-not-sufficient"-style statements	OK	unchanged; correct usage
RS5B codex packet	§3 point 6 "13 packet elements are conjunctive"	OK	inherits corrected item 5; no rewrite
RS5B rollup	§4 "effect_identity" bullet in the 13 elements	OK	inherits corrected item 5; no rewrite
RS5B-06	BI05 "approval … not bound to effect_identity / artifact_hash" → reject	OK	already enforces binding-to-effect; consistent with correction

3. Supersession ledger (authoritative)

superseded text (RS5B-05 item 5)	superseded by	effective wording
"the business-only effect digest (PATCH2 effect-identity) for the intended act, kept out of the authorization binding (separation)"	RS5B_PATCH1	PATCH1-02 §5 corrected item-5 text: effect_identity is pure (authority excluded) and is a required input of authorization_binding_digest

Nothing else in the RS5B package is superseded. RS5A-PATCH4 quorum/G02, the 13-element conjunction, item-13 Chairman primacy, the LEGO proof, and the verdict RS5B_READY_FOR_GPT_REVIEW are all unchanged.

4. Why no broader rewrite is warranted

The other three docs reference effect_identity only as a named element of the packet, never asserting it is "out of" the binding. Under the corrected contract those references are true (effect_identity is an element, and it is bound into the authorization digest). Rewriting them would be scope drift. The single contradiction is closed at its source (item 5) and the corrected reading propagates by reference.

5. Status

IMPACT_MAPPED — one direct SUPERSEDED phrase plus two minor shorthand corrections (RS5B-03 Model D and RS5B-05 item name); all other occurrences OK. All nine RS5B directory files plus the rollup were scanned. No package-wide rewrite; no RS5A reopening.