[derek73]

Closes #762. With #830, #831 and #832 merged, every decisions file except the five reference lists has a ## Standing block, so the scaffolding that held the rollout goes. This PR also fixes what round 3's reviews found.

The Standing rule, without its exempt list

• tests/v2/decisions_without_standing.json is deleted, along with STANDING_EXEMPT_ORIGINAL (the frozen original 88) and read_standing_exempt.

• REFERENCE_LISTS fixes the five reference lists, with its size asserted.

• standing_roster_findings(files, reference, allowed) flags:

  • a non-reference file with no block;
  • a reference list carrying a block, or shaped like a log;
  • a listed key outside the five;
  • a misspelled heading.

  Every finding the old check could still produce is kept. The four that were only about the list (gained a block, key with no file, key not in the original set, reference list also on the list) went with it. The live test is now test_every_decisions_file_opens_with_standing_or_is_a_reference_list.

• decisions/README.md and docs/design/AGENTS.md say so.

Round-3 leftovers

• rules/R3 (initials honor the fold): a base letter and its mark now initial whole only where they compose. A mark NFC leaves separate is left out of the initial: the nukta after ज, because U+095B is a composition exclusion. This is what decisions/R3's initials() drops the accent from a decomposed (NFD) name: émile zola gives e. z. #585 already decided, so it needed no issue.
• decisions/P8: the Should Jose E Maria Santos join like Jose e Maria Santos does? (the single-letter connective's Latin-capital veto was never adjudicated) #383 + JOSE E MARIA SANTOS reads the conjunction E as an initial; an all-caps name carries no case signal to read it from #479 entry's JUAN Y GARCIA "J. G." is marked superseded by rules.md#R3's "a conjunction never initials" is unqualified: Juan de y has base y and initials J. #461 (decisions/R3). Both views give "J. Y. G.".
• decisions/R1: the Keep citation comments short: quote, clause, pointer #760 entry's nee example is corrected (kind 2). Both inputs moved with Rethink the maiden-marker rule (M2): a marker stands after the current name and takes words up to the trailing suffix run #601, measured either side of 1037c87 under both policies. For Smith, … the comma was a family comma before as after.
• decisions/M2: Not standing: no longer lists Should a suffix delimiter inside a maiden clause count as a name word beside a link? Smith, John, PhD née Puig Mr. - i Soler keeps i Soler in the maiden name #538's resolution, which a later decision reversed. Its head note already carries that.
• decisions/parse-cost: a 2026-10-04 Rethink the maiden-marker rule (M2): a marker stands after the current name and takes words up to the trailing suffix run #601 entry records the link pin's re-pointing, 2,301 → 2,678. It is marked superseded by Share group's connective-join loop with the comma decision's title chain #617. The _CALL_BASELINE comment in test_benchmark names Where does Freiherr St John Smith MA start: should a title that is also a particle open the name? #624 and Merge the trailing run's particle runs in group, so S2's read needs no BOUND stop or report gathering #625.
• decisions/cjk-full-stops: the bullet approved 2026-09-09 gets its own dated line, and its two Standing lines point there.
• decisions/README.md: docs(decisions): Standing blocks for six named-topic files (#762) #832's 220-word bundle-dating sentence is now four sentences. The same-day rule is scoped to bullets whose text gives no date.

Review

Opus design-docs-reviewer, two rounds.

• Round 1, three findings. The split README sentence had widened the same-day rule. The new Rethink the maiden-marker rule (M2): a marker stands after the current name and takes words up to the trailing suffix run #601 entry presented 2,678 as current. R1's correction gave only the default policy.
• Round 2, clean. It re-checked every dated line in 3-0-reevaluations, comma-suffix-arc, cjk-full-stops and P2 against git log -S, and the old roster's findings one by one.

Mutation controls: switching off the no-block check fails 4 rows, and switching off the fixed-set check fails 1.

Tests

uv run pytest: 22,766 passed on 3.11 and on 3.15. Six fewer than before, because the exempt-list rows went. ruff and mypy are clean.

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.05%. Comparing base (d6a4d9f) to head (ee6029e).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #833   +/-   ##
=======================================
  Coverage   99.05%   99.05%           
=======================================
  Files          47       47           
  Lines        4659     4659           
=======================================
  Hits         4615     4615           
  Misses         44       44           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.