Law-gap sweep at revision 21: 9 CONFIRMED seams for the A20 draft to cure

Five independent sol lanes ran an adversarial law-gap sweep over the composition surface (§§19–33) at master 6e46e8a (revision 21, A13–A19 operative), each with a distinct lens drawn from one historical lawful-stop class. Consolidated report with full byte citations: ~/m6-sol-lanes/e8-ops/sol-ce-law-gap-sweep-r21-2026-08-16.md (lane transcripts noted per lane). Purpose: batch the next several lawful stops into one cure instead of paying for them serially at ceremony time — filed now because no A20 draft lane exists yet.

CONFIRMED (each would be a future lawful stop)

Data-reality / §19.3.3 class — live R04 blockers:

1. 46 of 818 complete-official prompts violate the singleton field-token rule. Witness: 1985 C68. resolves to both V11804 and V11805 → mandatory abort. A19's disposition law does not cure this class.

Supersession-map incompleteness (lane E, map ledger 34/34 rows reviewed):
2. A19 vs §§26.6.3/26.10.1 build order — A19 constructs purpose_mapping_rows before O_H (line 55784) while undisposed law requires O_H first (lines 50288, 51133); §33.7 disposes only the stale §19.3.3 anchor (line 56206).
3. §31.3 not joined to §28.2.1's iff-four-conditions operativity rule — the sole map link (line 55026) cites a nonexistent "A16 ratification demonstration" anchor. This is the seam class that caused the first repin stop; still open for every future amendment including A20.
4. §20.4.2's frozen Q5 era/proof shapes silently superseded twice by A19 (lines 55844, 55996; no covering §33.7 row) — first U == 0 build cannot satisfy both.
5. §25.6.6's literal interpreter token never disposed by A18 (line 48137 vs 55404; §32.7 calls the gate unchanged at 55620) — neither possible command[0] satisfies both contracts. Same defect class as the A18 CI-portability cure, one layer deeper.

Ordinal/pin class (lane A):
6. Terminal-A19 pinned fixtures repeat the A17 hardcoded-domain pattern at revision 22+.
7. A19-only implementation-pin resolution (breaks on the next amendment's pins).
8. R06's 223-test count is unbound to any pinned identity.

A20 charter conflict (lane D):
9. A20's chartered single-blob lifecycle law conflicts with §26.10.3 (lifecycle successors definable only post-certification). A20 as chartered is unratifiable without an express dormant-definitions supersession — otherwise lifecycle work needs an A21.

Key SUSPECTED (probes specified in the report)

• 15,428 prompts with zero field candidates — group by post-O_P positive row to test for empty required reference unions.
• MD= ceiling 54,898 vs 59,424 diagnostic shadow — historical/current representation bridge unproved.
• §33.8 manifest forbids serializing questionnaire_occurrence_rows that the global purpose rows appear to require (lines 56284 vs 27544).
• §31.3 verdict grammar: pinned validator silently requires LF + plain/comma decimals only (probed: space-grouped and CRLF forms abort) — enact an exact grammar.

Cleared (useful negatives)

§33.2 disposition totality over all 21,971 prompts (818/14/56/21,083); typed_value_unit lawfully resolvable via A10/§24 (sequencing question only); all active pins and both semantic-hash fixpoints verify at HEAD; A20 two-arm chronology satisfiable, incl. shared §31.3 receipt/verdict/operator-merge and the rev20→21→22→X→R05→R06 descendant order.

Recommendation

Fold cures 1–9 into the A20 draft (or an A20-pre batch amendment) before drafting starts — est. 3–4 avoided serial stop-and-cure ceremonies (~1–2 weeks of p50). Sweep findings are advisory; every claim carries byte citations for referee verification.

🤖 Generated with Claude Code

Start with the cited report at ~/m6-sol-lanes/e8-ops/sol-ce-law-gap-sweep-r21-2026-08-16.md, then verify its byte citations against the referenced §§19–33 and A19 anchors at master 6e46e8a. The work is done when cures for confirmed seams 1–9 are incorporated into the A20 draft or a pre-A20 amendment and the resulting claims are referee-verifiable.