Point the status page past the nineteenth prose review.

[unpaidmoon]

Change and mathematical scope

Navigation repair. No new κ. Selected conditional κ stays 0.001672660197076918.

docs/research/current-status.md called the nineteenth reconciliation the latest recorded round and left its disposition as "Review in progress". The nineteenth review already records 63 merged; 1 closed as superseded; 36 pending. Intake packages research/community-efficient2000-audit through research/community-efficient2600-audit are on this main. Round 2200's heading says "final unbatched remainder"; rounds 2300 through 2600 are recorded after it.

The status page now names the nineteenth reconciliation as the last prose row, copies that review's count, and links intake through the twenty-sixth reconciliation. The seven audit READMEs are not rewritten. There is no community-efficient2000-review.md through community-efficient2600-review.md, and this change does not invent those prose reviews.

This is not a supplier admission, a ledger rewrite, a certificate, or a selected-frontier change. It is a scoped correction of one historical-current sentence.

Task and overlap

Base: bb4bccbf42844d2801487450727c2f63e10d1d0f (canonical CrocSwap/integer-mult-bounds main). Task card scope-navigation, scope refined to that one sentence. Priority revision 6.

Donated session 25 took item 15 of the ordered eligible list after dropping claimed, delivered, and open-PR work through #4128. Items 1–10 stay with the in-flight sessions. The list used here:

1. T488 invoice-level ownership audit of composed PR Compose retimed T488 suppliers: conditional kappa 0.00173041889976 #3653 (Discussion Handoffs and requests for independent checks #1166 comment 19022532, question 1).
2. Physical rank-8/u2 source/consumer contract for 16 extra rank-one children (comment 19029639 / PR Certify integer-count budgets for three ideal lever combinations (no new kappa) #3889).
3. Independent review of research: certify local-first two-ADD obstruction on staged544 #3942 interaction/new-gap reduction (reply 19032637, task 1), distinct from Certify losing-first two-edit obstruction at displacement two on staged544 #3957, Certify neutral fixed-frame two-edit obstruction at displacement two #3970, Close stationary simultaneous-frame loophole on staged544 #3964, and Price one moved-ADD crossed-frame pair at histogram delta zero. #4122.
4. Adjacent completed-call pair, P104 two-axis gate (comment 19034066).
5. Integrate the A1858 scalar saving into the PR Raise conditional staged544 κ by filling idle owner slots #3945 producer (comment 19035159).
6. Source-normalization fusion caller (comment 19035513).
7. Physical producer/consumer half of the tensor-word handoff (comment 19035716). Audit the P104 index-to-tensor bridge without replaying events. #4125 covers the index-to-tensor bridge only.
8. Physical K/adjoint/frame price for native-lift tags (comment 19036235).
9. Price Context / source-vocabulary admission for the directory producer (comment 19037308).
10. Lower many-axis fixed-tape prime producer at the encoded caller boundary (comment 19037887).
11. q2 event 58: whole 76 versus stock 72, target local 16 (comment 19039339).
12. Catalogue card connector-alphabet-change.
13. Catalogue card formal-interface-lemma.
14. Catalogue card replay-runtime (title filter only; perf(research): compile the P112 ten-deletion census with popcnt (2.69x, byte-identical receipts) #4090 already did one lever).
15. Catalogue card scope-navigation — this PR.

Open PR titles through #4128 do not contain this card's overlap terms as a status-page repair. Cached open-PR bodies (821 records, fetched 2026-10-20) do not mention "latest recorded round", scope-navigation, or "twenty-sixth reconciliation". The nearest predecessor is merged #2892, which refreshed the same page through the nineteenth round and is released. This PR continues that navigation from 2000 through 2600.

Validation

CPython 3.12, from the repository root:

python3.12 -B scripts/research_package.py --package research/review-round-navigation
python3 -B -m unittest tests.test_research_current_status

Observed:

PASS_REVIEW_ROUND_NAVIGATION prose_last=1900 intake_last=2600 counts_1900=63_merged;_1_closed_as_superseded;_36_pending. selected=0.001672660197076918 rounds=7 controls=4 no_new_kappa
PASS_REGISTERED_RESEARCH_PACKAGE research/review-round-navigation outcome=completed

tests.test_research_current_status: 6 tests, OK. The historical-snapshot banner and its U+201D byte are unchanged.

The checker reads the working status page, the pinned nineteenth review, and the seven pinned intake READMEs. Four controls reject the old "latest recorded round" sentence, an in-progress nineteenth disposition, a changed twenty-sixth heading, and a prose review file for round 2600.

Not run: make frontier1094-verify, Lean, Docker, and a reread of every later intake row. PASS_REGISTERED_RESEARCH_PACKAGE is a navigation check, not mathematical acceptance.

Credit

The nineteenth reconciliation and the status-page structure through round 1900 are the existing review records, including the navigation refresh in #2892 that stopped at that round. Rounds 2000 through 2600 keep their authors' intake text. The sentence repair and the checker are by unpaidmoon, with substantial Cursor Grok assistance. Apache-2.0. No independent human review is claimed.

Open next questions

From research/review-round-navigation/README.md:

1. Should the prose table grow a row for each intake from 2000 through 2600? The count lines are already in the audit READMEs. A new row needs a community-efficientNNNN-review.md or an explicit decision that an intake README is the review record. Blocked until a maintainer chooses that record type.
2. Does docs/research/research-archive-intake.md still send publication state only through the nineteenth dispositions file? That page names research/community-efficient1900-audit/DISPOSITIONS.json. Later rounds have their own disposition files. First step: compare one later DISPOSITIONS.json with the ledger row for the same PR. Left open because editing that index was outside the one stale status sentence.

Workflow feedback (optional)

Overlap search against the GitHub search API returned HTTP 403, so the open-PR scan used the repository list API and scripts/find_overlap.py's cache. Feedback discussion: #1200