[unpaidmoon]

Change and mathematical scope

Scoped negative result. No new κ and no selected-frontier change. The selected record stays κ = 781451/250000000.

PR5679's ordinary component

C = 3507410738409/1000000000000000

cannot occupy the admitted bit_saving input of the selected PR6030 pricing. That pricing requires

atom > ordinary > admitted > TESTED

with ordinary recomputed from the thinned controller as 8590884599/2500000000000. Ordinary is strictly below C, so the gate fails. The outer ledger is not the rejecting check: its only bit_saving use is bit_saving − a, and every pinned check stays positive at bit_saving = C. The complex uniform saving 10940329/3500000000 stays below both 17/5000 and C, so accepting C as a ledger input would not raise κ.

C is strictly smaller than the PR5708 component 3507719059749/1000000000000000. Rejecting that larger value does not decide C. C still exceeds the selected ordinary stock. The cover that would make the ordinary formula equal C is 3507256898809/996000000000000, not the pinned 69/20000.

Task and overlap

Base commit 7c49fe32c55de02c2c873655c804b16e4c3fef0a on canonical CrocSwap/integer-mult-bounds main. Head c0915326ae206e89c1de73fc26b7f4668164f407. Priority revision 8.

Task card supplier-boundary-audit, scope status: refined. The card asks for one resolved admission gap on a latest candidate. The candidate is the PR5679 ordinary component. The resolved gap is the selected stock gate. Full native, caller, compiler, tape and analytic coverage stays open.

Session r114, item 7 of the ordered eligible list. Open handoffs on discussion #1166 were dropped when a later post completed them, a merged result answered them, the canonical artifact is an open pull request, or the post says no target was taken. That removes the delivered 19 October index, the H25 chain, and the 2 November review artifacts #6355–#6396. No structured claim was unexpired at the overlap refresh 2026-11-09T13:17:18Z. Three prose handoffs remain, the revision-8 next-decision table then supplies six rows, and unnamed catalogue cards follow:

1. Three signed added frames at seven calls (comment 19377832, updating 19377345).
2. Eight-source singleton connector at length 30 (comment 19377982).
3. Six-source depth-two block at c = 7 (comment 19384981).
4. Complex helpers (p104-side-correction, coverage-law-change).
5. BIT (bit-helper-contract).
6. New decoders and interfaces.
7. This package. Supplier audit (supplier-boundary-audit).

Nearest work:

• Record that the selected pricing rejects the PR5708 component as admi… #6343 rejects the larger PR5708 component at this same gate. It does not load the PR5679 receipt.
• Record that the G20 split input does not free a same-frame paid query. #6925 prices one paid-query identity on the Split one G20 helper into independent rank4 jobs at ordinary BIT saving 0.003507410738409 #5679 word and does not test this substitution.
• research/g20-independent-review recounts both components and leaves conversion into admitted BIT open.
• research/g20-global-caller-audit audits the PR5708 caller, not this receipt.

Overlap on 2026-11-09 searched 882 open PR bodies and 6,922 discussion records, plus local notes at 7c49fe32. The component decimal appears as PR5679's own saving. No open PR states this stock-gate rejection. No matches is not proof that a hidden diff is absent.

Validation

CPython 3.12, standard library only, from the repository root:

python3 -B research/g20-5679-caller-stock-gate/verify.py
python3 -O -B research/g20-5679-caller-stock-gate/verify.py
python3 -B scripts/research_package.py --package research/g20-5679-caller-stock-gate
python3 -B scripts/check_research_commit.py --package research/g20-5679-caller-stock-gate --check-source-history
python3 -B scripts/research_selected_baseline.py

Observed PASS_G20_5679_CALLER_STOCK_GATE, PASS_REGISTERED_RESEARCH_PACKAGE with outcome negative-result, PASS_RESEARCH_COMMIT_BYTES at c0915326ae206e89c1de73fc26b7f4668164f407, and PASS_RESEARCH_BASELINE_BINDINGS_ONLY. Checks raise SystemExit, so -O does not disable them.

Controls: bit_saving = 0.0027 still fails; bit_saving = a is not strict; dropping ε or using before_cover changes ordinary; the implied cover is not 69/20000 and is not the PR5708 cover.

Not run: the PR5679 caller replay, native maps, compiler, tape, analytic recovery, make frontier6030-verify, Lean, and Docker.

Credit

The component and the open caller question are Ali Romman's PR #5679, recorded in research/g20-helper-split. The selected pricing and claim are DreamingOfClouds' PR #6030. The ledger is the retained PR4359 endpoint. The gate pattern used for the larger PR5708 component is research/g20-component-stock-gate. Dugongue, ai-nanachi, and the notices retained in the G20 packages keep their credit.

Prepared by unpaidmoon with substantial Cursor Grok assistance. Apache-2.0.

Open next questions

1. Can a source-bound cover other than 69/20000 make the selected ordinary formula equal C? The implied value is 3507256898809/996000000000000. First step: read the thinned controller's cover field before constructing a cover. Blocked by a changed stock contract.
2. Can a Gaussian caller that does not use this pricing assembly consume C? First step: name that caller and its bit_saving input. Blocked by the absence of such a checked caller.

Workflow feedback (optional)

A link to discussion #1200 is enough; no separate session report.