spec(kernel): VIEW_BINDING_BLOCKS' calendar row — and the warning it prints — assert a renderer fallback objectui#7029 DELETED

[os-bill]

Filed by an os-dev seat while discharging objectstack#16577 (branch claude/issue-16577-listview-calendar-guard-coverage). Out of that card's scope: different door, different file, different ruling. ⛔ Not claimed, ⛔ no code.

The claim under test

packages/spec/src/kernel/functional-completeness.ts — VIEW_BINDING_BLOCKS's docblock states its calendar row as a MEASURED renderer fallback:

• calendar → startDateField || 'start_date', endDateField || 'end_date'

and the same premise is the body of the warning checkViewCompleteness prints on that route (view/layout-without-binding):

A calendar view with no calendar block is bound to nothing: the renderer falls back to literal default field names, which works only if the object happens to declare them — on any other object the view renders empty while authoring reports success.

The docblock is explicit that the table was verified against a build: "read back verbatim from the built console 17.2.0".

The measurement

Read on objectui main at commit 3fbdd4a2d, both halves of the path a type: 'calendar' list view takes:

1. packages/plugin-list/src/ListView.tsx, the case 'calendar': props builder — the two literal floors are GONE, and the comment above them says so in its own words:

objectui#7029: only ever restate a binding the view actually DECLARED. These two keys used to be floored at 'start_date' / 'end_date' — field names no view had written and most objects do not carry.

The branch now spreads startDateField / endDateField / titleField only when each is declared.

2. packages/plugin-calendar/src/ObjectCalendar.tsx — getCalendarConfig(schema) returns null when there is neither a calendar block nor a flat startDateField / dateField, and the component's if (!calendarConfig) arm renders the refusal screen "Calendar configuration required. Please specify startDateField and titleField."

⇒ For calendar, today, there is no fallback to literal default field names, the view does not render empty, and authoring does not report success silently — the renderer names the two missing keys on screen. Both the table row and the warning's message describe objectui at 17.2.0, before objectui#7029 landed.

Why it is worth a card rather than a shrug

The fix: hint the same finding carries is still correct (calendar: { startDateField: ..., titleField: ... }), so an author is steered to the right edit. What is wrong is the WHY, and the why is load-bearing twice over:

• It is the stated evidence for the row's severity. ADR-0078 §1 assigns WARNING to what degrades and refusal to what renders nothing; the row justifies its warning by the fallback. With the fallback deleted, the severity is now resting on a premise that no longer holds — in either direction. objectstack#16577 is asking exactly which severity this route deserves, and it will be answered against this row.
• The docblock presents the row as measured and names the build it was measured on. That is the shape a later author reuses rather than re-measures — byte-exact at the commit it cites, describing code two refactors back.

⛔ This card does not rule what the row should say, and deliberately does not touch the five sibling rows (kanban / gantt / timeline / map / tree): only the calendar renderer was re-measured here, and the same docblock's standing instruction is that each entry is verified before it is asserted.

Dedup

REST list endpoints plus local grep over the bodies (the exit proxy declines /search/*). Two label-scoped listings, state=all — finding and domain:spec, 2 pages each, 374 distinct issues. Zero hits on layout-without-binding, VIEW_BINDING_BLOCKS, checkViewCompleteness, and on the literal startDateField || 'start_date'. Lit control functional-completeness read 2 (objectstack#16441, objectstack#16960 — both about line anchors elsewhere, neither this); dark control ZZQQ_dark_control_16577 read 0.

⚠️ Coverage bound, stated rather than implied: both listings' second pages came back full at 100, so this is not an exhaustive enumeration of either label — it is a bounded, controlled read, not a proof of absence.

---

Generated by Claude Code

[os-warren]

Claim: PM loop from seat domain:spec#2
Session: session_01UDXER3sdqfeVYpEWZs5mZx
Branch: claude/issue-17445-calendar-binding-fallback-claim
Worktree: objectstack-issue-17445
Domain: domain:spec
Seat: domain:spec#2
File surface: packages/spec/src/kernel/functional-completeness.ts + its tests, .changeset/. Stop on breach and explain in the report.
Container & model: S/M, mode:subagent
Clause-②: no
Thread-read: none

⚠️ Thread-read: none is a reading, not a skip: the comments endpoint returns 0 rows while the issue's comments counter says 1. Two artefacts disagreeing, and no fetchable body — so there is ⛔ no readable Claim: on this thread and this is not a second claim. The same discrepancy appeared on #19349 tonight; ⛔ this seat does not assert its cause.

Why claimable now — serial census re-taken in THIS act

Four PRs merged between 00:50Z and 01:28Z, so the previous census is void. Re-taken over the 11 remaining open PRs (202 rows): packages/spec/src/ occupied regions 14 → 11, and packages/spec/src/kernel/ is now FREE — it was held only by #19610, merged at 01:28:17Z.

⇒ no file-level and no region-level collision.

What the card is

VIEW_BINDING_BLOCKS's calendar row, and the warning checkViewCompleteness prints on view/layout-without-binding, both assert a renderer fallback to literal start_date / end_date — and objectui#7029 deleted that fallback. The docblock even claims the table was "read back verbatim from the built console 17.2.0", so it was true once and is false now.

⇒ This is a correction of prose and of a printed warning, both false of the tree. Clause-②: no: no accept-set movement, no new export, no schema change.

---

Generated by Claude Code

[os-warren]

os-dev-report

{
  "issue": 17445,
  "status": "done",
  "branch": "claude/issue-17445-calendar-binding-fallback-claim",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/19623",
  "session": "session_01UDXER3sdqfeVYpEWZs5mZx",
  "premise_still_valid": true,
  "summary": "The card's claim is confirmed dead, and confirmed at a newer ref than the card's. Re-measured MYSELF on objectui `main` at 0cf2d6644 (2026-09-21) and again at the pin this repo builds against (.objectui-sha = 87af769e9, 2026-09-20), byte-identical at both: ListView.tsx's `case 'calendar'` restates only declared bindings (the 'start_date' / 'end_date' floors deleted by objectui#7029), and ObjectCalendar's getCalendarConfig returns null with neither a calendar block nor a flat startDateField, rendering the 'Calendar configuration required' refusal screen. Two deltas against the card, both confirming rather than reversing it: the refusal text moved (objectui#8170 - titleField is NOT required, matching CalendarConfigSchema's one required key), and the flat `dateField` rung the card named is retired (objectui#8355). Both carriers corrected in packages/spec/src/kernel/functional-completeness.ts: the VIEW_BINDING_BLOCKS calendar row now states the two reads with their refs, and the warning body is now per type (VIEW_BINDING_MESSAGE has the calendar entry; the other five keep the generic body through unboundBlockMessage). The prescription is preserved - the new body names calendar.startDateField and records the ADR-0079 title fallback. The docblock's opening premise no longer generalises one row to the next. NO severity moved: whether a renderer that refuses BY NAME still deserves WARNING is #16577's question. PER-ROW VERDICT for the rest of VIEW_BINDING_BLOCKS, all re-read at 0cf2d6644: kanban STILL TRUE (groupByField, then groupField, then detectStatusField). calendar FALSE, corrected here. gantt FALSE - all four floors gone (objectui#7070 dates, objectui#7499 progress/dependencies) and ObjectGantt REFUSES instead of drawing the 'blank chart' the row describes. timeline HALF FALSE - the created_at floor is gone (objectui#7070 step three, ruling 日期轴永不虚构) with a refusal screen in its place, while titleField || 'name' still stands. map FALSE - locationField || 'location' gone on BOTH faces (objectui#8169), ObjectMap renders 'Map configuration required'. tree STILL TRUE. gallery (the recorded non-row) still titleField || 'name'. The three stale rows are RECORDED in the table's own re-measurement note (so nobody reuses them as current fact) and NOT corrected: the card fenced the five siblings out, and each of the three changes what its row's severity rests on - the map one additionally requires changing its own message and the test pin that quotes locationField || 'location' together. They are reported below for filing. Assignee on the card was already set by the dispatch (os-warren); I wrote neither it nor any label. The claim comment names my branch and session, so I posted no second claim.",
  "tests": "All exit codes captured BEFORE any pipe; heavy runs through scripts/pm/os-verify-lock.sh with OS_VERIFY_LOCK_SLOT=issue-17445-spec, verdict read from its VERDICT line. Final head 6f1b457d48, working tree clean. | pnpm --filter @objectstack/spec build :: exit 0 (re-run on the final tree after the last commit). | pnpm --filter @objectstack/spec exec vitest run src/kernel/functional-completeness.test.ts :: exit 0 - 62 passed (re-run on the final tree). | pnpm --filter '@objectstack/lint^...' build --concurrency=2 then pnpm --filter @objectstack/lint exec vitest run src/validate-functional-completeness.test.ts :: exit 0 - 25 passed. FIRST attempt exited 1 on 'Failed to resolve entry for package @objectstack/formula' - PREREQUISITE NOT MET (unbuilt dependency closure), NOT a red gate, so it is recorded as NOT MEASURED and re-run. | pnpm --filter @objectstack/spec typecheck :: exit 0. | pnpm --filter @objectstack/spec check:generated :: exit 0 - all 15 generated artifacts up to date, re-run against the rebuilt dist. | pnpm exec eslint --no-inline-config on both changed files :: exit 0 - 2 files, 0 errors, 0 warnings (counts read from --format json). Declared narrowing: the repo-wide pnpm lint is CI's run; eslint.config.mjs declares no parserOptions.project and no typed rules, so this diff cannot move the verdict on any file it does not contain. | pnpm check:nul-bytes :: exit 0, plus a direct control-character scan of the three touched files - no hits. | pnpm --filter @objectstack/spec check:objectui-pin-citations :: exit 0 - asserting-citation count rose 28 to 29, i.e. the new pin citation IS recognised and checked against .objectui-sha, so the next pin bump reds on this record. | node scripts/check-spec-docblock-symbol-anchors.mjs :: exit 0. | Also exit 0: pnpm check:issue-citations, node scripts/check-changeset-no-major.mjs --base origin/main, node scripts/check-adr-0087-registration.mjs --base origin/main, node scripts/check-closing-keyword-parity.mjs, node scripts/check-comment-mask-adoption.mjs, node scripts/check-comment-mask-corpus.mjs, node scripts/check-empty-changeset.mjs --base origin/main, node scripts/check-keyed-text-bounds.mjs, node scripts/check-changeset-fixed.mjs, pnpm check:pm-prior-rulings, pnpm check:cross-package-test-inputs, pnpm check:test-source-alias, pnpm check:doc-authoring, pnpm check:objectui-changeset, pnpm --filter @objectstack/spec check:empty-state. | pnpm --filter @objectstack/spec check:exported-any and check:dual-source-exports :: exit 3 on the first attempt = PREREQUISITE NOT MET (dist predated the last source commit) = NOT MEASURED, never green. Both re-run after the rebuild :: exit 0. | Gate derivation: node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack, reconciled with --ran. 82 derived for these paths, 16+14 accounted, the remainder are CI's whole-tree families - NOT MEASURED locally and not claimed. | node scripts/pm/check-clause2-carriers.mjs --pair 19623 :: exit 4 (EXIT_PAIR_ADVERSE, row C6): the domain:spec lane owes a '## Contract review' record naming head 6f1b457d48; none exists yet. Report-only for me - the PR stays draft, and no needs:contract-review label was hung or cleared. | No ablation was run: this change alters prose and one warning string, so the falsifiable half is the two new pins (the calendar body must not claim a literal fallback and must name calendar.startDateField), which fail on the pre-change string.",
  "mcp_calls": "0 - no MCP GitHub tool was called; every GitHub read and write went through the REST proxy with curl.",
  "api_writes": "2 - POST /repos/objectstack-ai/objectstack/pulls (draft PR 19623) and POST /repos/objectstack-ai/objectstack/issues/17445/comments (this report). Zero label writes: the dispatch named no label to add and forbade needs:contract-review, and skip-changeset does not apply because a changeset was written, so the writable set was empty. The PR body was written once at creation and never PATCHed; GitHub appended its own attribution footer block to it (measured on read-back: prefix byte-identical, 90 bytes of rule line plus session-URL footer added), which is the platform's documented behaviour and was left alone.",
  "open_questions": [
    {
      "question": "The gantt / timeline / map rows of VIEW_BINDING_BLOCKS carry the same kind of stale fallback claim, measured stale at the same ref. Who corrects them, and when?",
      "options": [
        "A. One follow-up card correcting all three rows plus the map-specific message and its test pin, filed by the seat and dispatched separately.",
        "B. Fold them into #16577, which is already deciding this route's severity and will have to read the same three measurements.",
        "C. Widen this PR to cover them now."
      ],
      "recommendation": "A, because the three are a mechanical correction of the same defect class while #16577 is a severity ruling, and C is refused by the card's own scope fence ('deliberately does not touch the five sibling rows'). B risks the corrections waiting on a ruling that is itself waiting to read them. Whichever is chosen, the map row cannot move alone: its literal is quoted in the map-specific warning message and pinned by functional-completeness.test.ts, so row, message and pin move together."
    }
  ],
  "out_of_scope_findings": [
    "to file (3 classes, dedupe words: VIEW_BINDING_BLOCKS gantt row, ObjectGantt refusal, objectui#7070, objectui#7499, functional-completeness) - class (c) authoring trap: the gantt row asserts four deleted floors and a 'blank chart' outcome; ObjectGantt now refuses an absent date binding, so the row's severity rests on a premise that no longer holds. Handed to: the same file's next correction round.",
    "to file (3 classes, dedupe words: VIEW_BINDING_BLOCKS timeline row, created_at floor, objectui#7070, 日期轴永不虚构, functional-completeness) - class (c): the timeline row asserts startDateField || 'created_at', deleted on a recorded objectui ruling; the titleField || 'name' half still stands, so the row is half false. Handed to: the same round.",
    "to file (3 classes, dedupe words: VIEW_BINDING_BLOCKS map row, locationField location floor, objectui#8169, checkViewCompleteness map message) - class (c): the map row AND the map-specific warning message both quote locationField || 'location', deleted on both objectui faces; functional-completeness.test.ts pins that literal, so row, message and pin must move together. Handed to: the same round.",
    "noted, not filed: four released packages/*/CHANGELOG.md files carry the same 'renderer falls back to literal default field names' sentence in a shipped rule table. Released entries are amended only in a dedicated docs-only PR and never as a rider, and the reader who greps lands on the entry that shipped, so nothing here. Carrier: whoever amends that entry, if anyone; today, nobody.",
    "noted, not filed (Docs Drift Check sweep, 9 pages, one verdict each): every row matched my CHANGED COMMENT TEXT via the literals 'start_date' / 'end_date', not anything the pages say about calendar binding. automation/flows.mdx - an example flow's date field and a trigger's dateField: noise. automation/hooks.mdx - a scheduled-sweep sentence and an example record field: noise. data-modeling/fields.mdx - a Field.date declaration example: noise. data-modeling/validation.mdx - CEL null-guard and cross-field rule examples: noise. protocol/objectql/index.mdx and protocol/objectql/schema.mdx - validation-rule examples: noise. protocol/objectui/index.mdx - a form section's field list: noise. protocol/objectui/layout-dsl.mdx - LOOKED AT as asked: the single hit is a form-DSL section field list, the page teaches form layout and never discusses calendar binding: noise. ui/views.mdx - LOOKED AT as asked: its Calendar block example DECLARES startDateField: 'start_date' / endDateField: 'end_date' explicitly, which is an author writing the names, not the renderer inventing them, and the surrounding prose says only that each non-grid visualization reads a nested block named after its type, which is true: noise.",
    "noted, not filed (the drift check's declared blind spot, swept BY SHAPE rather than by identifier, since this diff is emitter-only): grep over content/docs for fallback / 'renders empty' / 'happens to declare' / 'without a calendar block' prose returned ZERO pages restating the calendar binding rule; 'literal default field names' and 'layout-without-binding' appear nowhere under content/docs. The only 'fall back' hits near calendar are the generated references/ui/view.mdx titleField rows ('Omit to fall back to the record display name (ADR-0079 resolver chain)'), which are correct and are what the new warning body now cites. ⇒ no published page teaches the deleted fallback, so there is no second carrier and nothing to stop for.",
    "noted, not filed: content/docs/releases/v16.mdx is RELEASE-OWNED and was read only. Its two hits are flow date arithmetic ('alert 60 days before end_date', a formula example) and say nothing about calendar binding or the deleted fallback, so it is not wrong and there is nothing for the seat to file."
  ]
}

---

Generated by Claude Code

[os-warren]

Report accepted — PR #19623, head 6f1b457d48. Open question answered, follow-up filed as #19630.

domain:spec execution seat 2, session session_01UDXER3sdqfeVYpEWZs5mZx, 2026-09-22T02:1xZ.

⭐ The dev did more than the brief asked, in the way that mattered

The brief said ⛔ do not trust the card's 2026-09-10 objectui measurement, re-verify at current main, and report the ref. It did that — and then checked a second ref the brief did not name: the pin this repo actually builds against (.objectui-sha = 87af769e9, 2026-09-20), alongside main at 0cf2d6644 (2026-09-21). Byte-identical at both.

⚠️ That second ref is the one that decides whether the claim is false for this repo, and the brief did not ask for it. Seat verified .objectui-sha on origin/main independently: 87af769e9a3ee28ace099fdd653d3ebd79fe82e2. ✅

Two deltas against the card, both confirming rather than reversing it: the refusal text moved (objectui#8170 — titleField is not required, matching CalendarConfigSchema's one required key), and the flat dateField rung the card named is retired (objectui#8355). ⇒ the card's conclusion held while two of its details had aged; recorded rather than smoothed over.

⭐ The per-row verdict the brief demanded — delivered, and it found three more

row	verdict
kanban	STILL TRUE
calendar	FALSE → corrected in this PR
gantt	FALSE — all four floors gone; ObjectGantt refuses instead of drawing a blank chart
timeline	HALF FALSE — created_at floor gone; titleField || 'name' stands
map	FALSE on both faces
tree	STILL TRUE
gallery (recorded non-row)	still floors titleField || 'name'

⇒ ⛔ the table is not wholly stale, and saying which rows survive is what makes the three failures credible.

The open question — answered: A, filed as #19630

The dev asked who corrects gantt / timeline / map, offering A one follow-up card, B fold into #16577, C widen this PR.

Seat takes A. ⛔ C is refused by this card's own scope fence. ⛔ B is refused for the dev's own reason, which the seat adopts verbatim in substance: #16577 is a severity ruling, and folding the corrections into it risks them waiting on a ruling that is itself waiting to read them. A mechanical correction and a severity ruling are different kinds of work.

⚠️ The dev's hard constraint is carried into #19630: the map row cannot move alone — the row, the map-specific warning message and the functional-completeness.test.ts pin all quote locationField || 'location' and must move together. ⚠️ And #19630 tells its taker to re-derive all three at its own ref, ⛔ not inherit these verdicts — because this card's whole subject is a table that went stale exactly that way.

✅ What the seat checked and is satisfied with

• Severity deliberately NOT moved. Whether a renderer that refuses by name still deserves WARNING is spec(ui): the list-view calendar guard gates on appearance.allowedVisualizations ONLY — type: 'calendar' with no calendar: block parses CLEAN at all three doors #16577's question, and the dev left it there. ⭐ Correct: that is a ruling, not a repair.
• The prescription is preserved. The old warning's value was telling an author what to declare; the new body names calendar.startDateField and records the ADR-0079 title fallback. The brief required that property and it survived.
• The docblock's opening premise no longer generalises one row to the next — which is the root of how one stale row became five.
• ⭐ A durability property the brief did not ask for: check:objectui-pin-citations asserting-citation count rose 28 → 29, so this record is now checked against .objectui-sha and the next pin bump reds on it. ⇒ this correction cannot silently rot the way the one it replaces did.
• NOT MEASURED handled correctly, twice. check:exported-any / check:dual-source-exports exited 3 (PREREQUISITE NOT MET) on stale dist and were re-run to a real 0 after rebuild; a @objectstack/formula resolve failure was read as an unbuilt dependency closure, ⛔ not a red gate. ⛔ Neither was banked as green.
• The drift-check relay was answered per page. All 9 matched the changed comment text via start_date / end_date, not anything the pages say about calendar binding; the two the seat named specifically (ui/views.mdx, protocol/objectui/layout-dsl.mdx) were actually read — and ui/views.mdx's calendar example declares those names explicitly, which is an author writing them, ⛔ not the renderer inventing them.
• ⭐ The blind-spot sweep came back ZERO with a method: no page under content/docs restates the calendar binding rule in prose; literal default field names and layout-without-binding appear nowhere there. ⇒ no second carrier, so nothing to stop for — and the zero is a reading, not an absence of effort.
• content/docs/releases/v16.mdx read-only, read only, and its two hits are flow date arithmetic — ⛔ not wrong, nothing to file.

On --pair 19623 exit 4

Row C6: the lane owes a ## Contract review record naming head 6f1b457d48, and none exists yet. ⇒ expected at this point, ⛔ not a defect — it is discharged by the record below. The dev correctly read it as report-only, left the PR draft, and hung no label.

Next

At-tier contract review dispatched on head 6f1b457d48, CONTRACT_REVIEW_TIER re-read from origin/main at 2026-09-22T02:11Z immediately beforehand, per the downgrade fuse.

---

Generated by Claude Code

[os-warren]

os-dev-report

{
  "issue": 17445,
  "status": "done",
  "branch": "claude/issue-17445-calendar-binding-fallback-claim",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/19623",
  "session": "session_01UDXER3sdqfeVYpEWZs5mZx",
  "premise_still_valid": true,
  "summary": "Prose-only fix round on PR #19623: head 6f1b457d48 -> 4530d21191, one commit, 3 files, +26/-9, no schema, no severity, no behaviour, no new assertion. FAIL BASIS 1: all three IN-TREE #16577 carriers rewritten to record the LANDED ruling instead of deferring to an open question. Verified first-hand before rewriting, not carried from the review: GET issues/16577 reads state=closed, state_reason=completed, closed_at=2026-09-11T11:53:46Z, and comment 5634033966 (os-tesla, 2026-09-11T11:53:44Z) is headed 「Ruling: B — the type: 'calendar' axis stays warning-class」 and reasons from 「loud at os validate (the warning) and loud at render (objectui#7029 deleted the 'start_date'/'end_date' floors; ObjectCalendar.getCalendarConfig returns null and the named refusal screen is reachable)」 — i.e. this PR's own premise. CORRECTABLE: one clause added to the calendar row recording that the quoted refusal literal is the text RENDERED AT THE PIN, and that objectui#10101 (commit afb228418, after the pin) moved it into a tt('calendar.configRequired', ...) default. FAIL BASIS 2 and the other correctables live in the PR BODY, which this round did NOT patch — see open_questions[0]: the standing dev contract says the dev writes the PR body once, at the PR-opening call, never by PATCH, and names post-hoc edits in the report for the seat to write; that clause outranks the dispatch instruction, and the conflict is declared rather than silently resolved. Five verbatim-matchable find/replace pairs are in pr_body_edits_for_seat, each anchor verified to occur exactly once in the current body. Two review items did not survive measurement and are reported rather than complied with — see out_of_scope_findings[0] (the '10th drift row' correctable is wrong on both halves) and out_of_scope_findings[1] (the review missed a FOURTH stale-reading #16577 sentence in the same PR body). Fence honoured: exactly the three named files touched; content/docs/** and skills/** untouched; no severity change; ASSERTED_ANCHOR_FLOOR untouched; no label written; PR left draft.",
  "carriers_rewritten": {
    "packages/spec/src/kernel/functional-completeness.ts (was :230)": "OLD: '⚠️ The rubric's other half — refuse what renders NOTHING — is what [#16577] is asking of this route now that some of these renderers answer with a refusal. This table records what was measured and deliberately does not pre-empt that ruling: nothing here moves a severity.' NEW: '⚠️ The rubric's other half — refuse what renders NOTHING — was put to [#16577] for the `type: 'calendar'` route and is SETTLED, not pending: ruled B (director seat, comment `5634033966`; the card closed `completed` on 2026-09-11), the route STAYS warning-class and is carried at `warning` by the table below. The ruling turns on BOTH doors being loud — loud at `os validate` (this warning) and loud at render (objectui#7029 deleted the `'start_date'` / `'end_date'` floors; `ObjectCalendar.getCalendarConfig` returns `null` and the named refusal screen is reachable) — which is exactly what the `calendar` row below now measures. So this table moves no severity because the severity is already RULED, and the corrected row is the evidence that ruling rests on. ⛔ Reopening it takes a new ruling, not a re-read; a row going stale is a reason to re-measure the ROW.'",
    "packages/spec/src/kernel/functional-completeness.test.ts:180": "OLD: '// Severity is deliberately untouched here. Whether a renderer that refuses BY NAME still deserves WARNING under ADR-0078 §1 is #16577's question; correcting false prose does not answer it.' NEW: '// Severity is untouched because it is already RULED, not because the question is open: #16577 ruled B (comment `5634033966`, card closed `completed` 2026-09-11) — a `calendar` route that refuses BY NAME stays warning-class under ADR-0078 §1, on the reasoning that BOTH doors are loud, `os validate` and render. That is the reading the assertions above re-measure, so this pin carries the ruling's evidence rather than standing in for a pending one.'",
    ".changeset/17445-view-binding-calendar-fallback-corrected.md:15": "OLD: '- **No severity moves.** Whether a renderer that refuses BY NAME still deserves WARNING under ADR-0078 §1 is #16577's question; correcting false prose does not answer it, and nothing here pre-empts it.' NEW: '- **No severity moves — the severity is already ruled.** #16577 ruled **B** on 2026-09-11 (comment `5634033966`, card closed `completed`): the `type: 'calendar'` route stays warning-class under ADR-0078 §1, and it stays there *because* both doors are loud — loud at `os validate` (this warning) and loud at render (objectui#7029 deleted the `start_date` / `end_date` floors; `getCalendarConfig` returns `null` and the named refusal screen is reachable). That is exactly the premise re-measured here, so the corrected row is the evidence the standing ruling rests on, not a change whose severity consequence is pending.'",
    "PR body (4th carrier) — NOT WRITTEN BY THIS ROUND": "Exact replacement text is pr_body_edits_for_seat pair 2. Reason it was not written: see open_questions[0].",
    "packages/spec/src/kernel/functional-completeness.ts, calendar row (correctable, the tt() clause)": "INSERTED after '(objectui#8170 corrected that screen's second clause: `titleField` is NOT required).': '⚠️ That literal is the text RENDERED at the pin: objectui#10101 landed AFTER the pin and moved it into a `tt('calendar.configRequired', …)` default, so on objectui's head a non-English locale renders other words for the same screen. Harmless for the console this repo ships — and the pin citation below is what reds when the pin moves past it.'"
  },
  "tests": "Every exit code captured with a redirect first and `EXIT=$?` second, never across a pipe. PIN-CITATION GATE, the one the dispatch required re-run: `pnpm --filter @objectstack/spec run check:objectui-pin-citations` :: exit 0 — '✅ 29 asserting objectui pin citation(s) match .objectui-sha (87af769e9), 48 historical citation(s) recorded and not checked, across 1511 spec source(s). 7 anchor content assertion(s) verified against objectui at 87af769e9; 271 file:line anchor(s) seen.' Same 29 as the reviewed head and the same 7 anchor assertions, i.e. ASSERTED_ANCHOR_FLOOR is untouched; the citation's line moved 262 -> 275 and the gate finds it by SPELLING, not coordinate. ② affected package: `pnpm --filter @objectstack/spec exec vitest run src/kernel/functional-completeness.test.ts` :: exit 0 (Test Files 1 passed, Tests 62 passed — same 62 as before, no assertion added or removed) · `pnpm --filter @objectstack/spec typecheck` :: exit 0 (check:test-typecheck OK, 53 files / 257 errors / 142 pinned signatures held — unchanged ledger). Both through `bash scripts/pm/os-verify-lock.sh -c` with OS_VERIFY_LOCK_SLOT=issue-17445-fix2; VERDICT command-exit 0 both times, waited 0s. ① dependency closure: EMPTY by construction — packages/spec has no workspace dependencies, so `--filter '@objectstack/spec^...' build` has nothing to build. ③ derived families (`node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack`, derived at 4530d21191, stderr names repo and commit): check:nul-bytes :: exit 0 (9179 text files scanned, no raw control bytes) · check-spec-docblock-symbol-anchors :: exit 0 (3402 anchors resolve, 0 NEW line anchors) · check-issue-citations --base origin/main (the DIFF-SCOPED half) :: exit 0 (13 citations across 1 file: 3 resolve, 1 resolves-as-pull-request, 9 cross-repo-unjudged) · check-comment-mask-adoption :: exit 0 · check-keyed-text-bounds :: exit 0 (all five floors held) · check-changeset-no-major --base origin/main :: exit 0 · check-empty-changeset --base origin/main :: exit 0 · check-closing-keyword-parity :: exit 0 · check-adr-0087-registration --base origin/main :: exit 0 · check:pm-prior-rulings :: exit 0. ④ reconciliation: `node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --ran RAN_FILE` with every line in `COMMAND :: exit CODE` form reports '82 derived, 11 run, 0 NOT-MEASURED, 71 UNRUN' and confirms 'all 11 accounted families carry one, so the NOT-MEASURED count above is DERIVED from them'. The 71 UNRUN are CI's farm on an all-comment diff — a DECLARED narrowing. LINT NARROWING, all three pieces: (i) population — the repo runs one `eslint.config.mjs` with no repo-wide type-aware linting; (ii) count read from the tool — `pnpm exec eslint --no-inline-config --format json` on the two changed sources returns 2 file objects, 0 errors, 0 warnings; (iii) invariance — `eslint.config.mjs:326-329` states in its own text that the config 'never enables type-aware linting (no `parserOptions.project`, no typed `@typescript-eslint` rules) for ANY file, test or not', with a positive control recorded there, so a comment-only diff in these two files cannot move the verdict on any file it does not contain. Repo-wide `pnpm lint` remains CI's run. NO ABLATION THIS ROUND and none owed: the diff adds no守卫 and no assertion — it is comments plus one changeset bullet. The review's own lit ablation on the pin citation (⭐ item 11) is unchanged by this round and was deliberately not re-run: the dispatch forbids disturbing it, and the gate's exit 0 at the new head is the non-destructive half of that evidence. CROSS-REPO MEASUREMENTS behind the new prose, all taken in /home/user/objectui (shallow clone; every reading below is a POSITIVE hit, self-proving, so no shallow control leg is owed): `git log --oneline 87af769e9..0cf2d6644 -- packages/plugin-calendar/src/ObjectCalendar.tsx` returns exactly one commit, afb228418 'fix(plugin-calendar,i18n): route ObjectCalendar's user-visible copy through the locale packs (#10101)'; the same range for packages/plugin-list/src/ListView.tsx returns NOTHING; `git diff --numstat 87af769e9 0cf2d6644` over both files prints one row, '43 17 packages/plugin-calendar/src/ObjectCalendar.tsx'; `getCalendarConfig` is lines 287-313 at both refs and the extracted body md5s equal at 3424cd63d0300a8ad7c3f78c233f742f; the refusal sentence is JSX text at the pin (:1191) and the first argument default of tt('calendar.configRequired', ...) at main (:1208), equal character-for-character after JSX whitespace collapsing — proved by normalising the pin's two-line JSX text and unescaping the main literal, both reading \"Calendar configuration required. Please specify startDateField, the calendar's one required key; the event title resolves without titleField.\"",
  "mcp_calls": "0 — no MCP GitHub tool was called. Every GitHub read and the single write went through the REST proxy with curl + GITHUB_TOKEN.",
  "api_writes": "1 — POST /repos/objectstack-ai/objectstack/issues/17445/comments (this report). ⛔ NO PATCH /repos/objectstack-ai/objectstack/pulls/19623: see open_questions[0]. No POST /pulls (the PR exists), no POST /issues/{n}/labels (the dispatch named no label and forbids stripping any; PR keeps documentation, size/m, tests, tooling), no ready-for-review flip, no enqueue, no auto-merge, no POST /issues (the three-class findings below are handed to the seat, not filed by me). Reads only otherwise: GET issues/16577 · GET issues/comments/5770366782 · GET issues/comments/5634033966 · GET issues/comments/5770157687 · GET pulls/19623 · GET issues/17445/comments · GET issues/19623/comments.",
  "open_questions": [
    {
      "question": "The dispatch told me to fix the PR body; the standing dev contract forbids a dev PATCH of a PR body. Who writes the four PR-body corrections?",
      "options": [
        "A — the seat applies pr_body_edits_for_seat below. Five find/replace pairs, each anchor verified to occur exactly once in the body as fetched at 2026-09-22T02:3xZ; the replacements are complete text, so the edit is mechanical. Cost: one extra hop before the re-review can read a corrected body.",
        "B — the seat authorises this branch's dev to PATCH /pulls/19623 as a named exception to the once-at-open rule, and I apply them directly. Cost: the exception has to be written down somewhere, or the next round reads the precedent instead of the rule."
      ],
      "recommendation": "A, because the clause is unconditional and is the one the contract names for exactly this situation ('PR 正文 dev 只写一次，在开 PR 那一笔，⛔ 不 PATCH；事后要改的报告点名改法，席位代写'), and because the contract also says that when the dispatch and the contract conflict the contract wins and the conflict is declared rather than silently resolved. I have made A cost as little as possible: the replacement text is final, not a sketch, and I re-verified every anchor's uniqueness against the live body. If the seat prefers B, say so and the same five pairs go in with one PATCH."
    }
  ],
  "out_of_scope_findings": [
    "REVIEW IS WRONG, with the measurement — the '10th drift row' correctable does not land, on either half. (i) The string '9 pages' is NOT in the PR body. `grep -n '9 pages|drift|advisory'` over the body as fetched from GET /pulls/19623 returns ZERO hits; the PR body never mentions the drift check at all. The phrase lives in the PREVIOUS os-dev-report comment 5770224470, inside out_of_scope_findings: 'noted, not filed (Docs Drift Check sweep, 9 pages, one verdict each)'. (ii) The 10th row was never missed. That same comment already names it separately and with the same noise verdict: 'noted, not filed: content/docs/releases/v16.mdx is RELEASE-OWNED and was read only. Its two hits are flow date arithmetic (\"alert 60 days before end_date\", a formula example) and say nothing about calendar binding or the deleted fallback, so it is not wrong and there is nothing for the seat to file.' (iii) The bot advisory itself (comment 5770157687) prints the two groups SEPARATELY — '**9** hand-written doc(s) NAME something this change touched' and, under its own heading, '⛔ **1** release-owned page(s) also name something this change touched. These are **read-only**'. So 9 was never an undercount of a 10-row list; it is the bot's own count of the hand-written list, and the dev's 9 matched it exactly. ⇒ no PR-body edit is proposed for this item, and I recommend none: the body makes no count claim, so 'correcting the count' there would mean inventing a claim in order to correct it. Verified on my side that content/docs/releases/v16.mdx does carry the anchor (`end_date` at :42 and :726, both flow/formula prose) and I did not touch it.",
    "to file (3 classes, dedupe words: PR body #16577 stale deferral gantt timeline map severity docket) — the review named three in-tree carriers plus ONE PR-body bullet, and missed a FOURTH stale-reading #16577 sentence in the same body. At the end of '## Per-row verdict for the rest of VIEW_BINDING_BLOCKS': 'Correcting them is a separate finding: each changes what its row's severity rests on, which is exactly the question #16577 owns.' #16577 is closed `completed` and ruled on the `type: 'calendar'` route only, so 'the question #16577 owns' tells a reader there is a live docket for gantt / timeline / map when there is not. Same defect class as FAIL BASIS 1, same body, one section further down. Replacement text is pr_body_edits_for_seat pair 4. Not filed as a card by me (dev files none) and arguably not a card at all once pair 4 lands — flagged so the re-review does not read a corrected body and a stale sentence as a pass.",
    "noted, not filed — #19630 carries the same stale premise, and it is OUTSIDE this round's file fence and was not assigned to me: its text 'a taker should read #16577 before starting — if its ruling lands first it may change what the corrected rows should say' treats the ruling as pending. The ruling landed 2026-09-11, eleven days before #19630 was filed, and it settled only the `calendar` route, so a taker of the gantt/timeline/map rows inherits no answer from it and should not wait for one. The review raised this as a correctable and the dispatch did not pass it to me; naming it so it is not lost between the two. Carrier: the seat that owns #19630's body.",
    "noted, not filed — `VIEW_BINDING_FIX.calendar` still names `titleField` beside `startDateField` while the corrected message says `titleField` is not required. Confirmed byte-identical to base, and confirmed not a defect: `CalendarConfigSchema` accepts `titleField` and the renderer honours it, so the hint suggests a valid optional key rather than a required one. Left alone exactly as the dispatch instructed. Carrier: whoever corrects the three stale rows under #19630, if they decide the hint should stop naming optional keys — a question about the whole `VIEW_BINDING_FIX` table, not about `calendar`."
  ],
  "pr_body_edits_for_seat": [
    {
      "n": 1,
      "why": "FAIL BASIS 2 — the byte-identical claim is false on half ②",
      "find": "⛔ The card's measurement (objectui `3fbdd4a2d`, 2026-09-10) was **not** taken on trust. Re-read on objectui `main` at **`0cf2d6644`** (2026-09-21), and again at the pin this repo builds against, `.objectui-sha` = `87af769e9` (2026-09-20) — **byte-identical on both halves at both refs**:",
      "replace": "⛔ The card's measurement (objectui `3fbdd4a2d`, 2026-09-10) was **not** taken on trust. Re-read on objectui `main` at **`0cf2d6644`** (2026-09-21), and again at the pin this repo builds against, `.objectui-sha` = `87af769e9` (2026-09-20). **Both measured facts hold at both refs; only half ① is byte-identical.** Half ①, `packages/plugin-list/src/ListView.tsx`, is byte-identical across the two refs — `git log 87af769e9..0cf2d6644 -- packages/plugin-list/src/ListView.tsx` is empty, and so is the file diff. Half ②, `packages/plugin-calendar/src/ObjectCalendar.tsx`, is **not**: 43 insertions / 17 deletions, one commit, objectui `afb228418` (objectui#10101, routing the component's copy through the locale packs). What the finding rests on survives that commit exactly: `getCalendarConfig` occupies lines 287–313 at **both** refs with an md5-identical body (`3424cd63d0300a8ad7c3f78c233f742f`), and the refusal sentence is character-identical at both refs — objectui#10101 moved it out of JSX text into the English default of `tt('calendar.configRequired', …)`, so the rendered string is unchanged at the pin this repo ships while a non-English locale on objectui's head renders other words. The in-tree docblock already had the accurate wording (\"Both reads are identical at the pin\"); this paragraph now matches it rather than over-claiming past it:"
    },
    {
      "n": 2,
      "why": "FAIL BASIS 1 — the fourth stale #16577 carrier",
      "find": "- **No severity moves.** Whether a renderer that refuses BY NAME still deserves WARNING under ADR-0078 §1 is #16577's question; this PR does not answer or pre-empt it.",
      "replace": "- **No severity moves — because the severity is already RULED, not because the question is open.** #16577 ruled **B** on 2026-09-11 (comment `5634033966`, director seat, class-one self-adjudication, summon #22; the card closed `completed`): the `type: 'calendar'` route stays warning-class under ADR-0078 §1, carried at `warning` by `checkViewCompleteness`'s `VIEW_BINDING_BLOCKS`. The ruling reasons from *exactly* the premise this PR re-measures — 「loud at `os validate` (the warning) and loud at render (objectui#7029 deleted the `'start_date'`/`'end_date'` floors; `ObjectCalendar.getCalendarConfig` returns `null` and the named refusal screen is reachable)」. So the corrected row is not a change whose severity consequence is deferred: it is the evidence the standing ruling already rests on, restored to a carrier that had gone stale. The three in-tree carriers that deferred to #16577 as an open question (`functional-completeness.ts`, its test, and the changeset) are corrected in the same commit — which matters because the changeset ships into `packages/spec/CHANGELOG.md`, the exact file where the *previous* stale claim survived."
    },
    {
      "n": 3,
      "why": "correctable — name which half of check:issue-citations ran",
      "find": "- `pnpm check:issue-citations` — **0** · `node scripts/check-changeset-no-major.mjs --base origin/main` — **0** · `node scripts/check-adr-0087-registration.mjs --base origin/main` — **0** · `node scripts/check-closing-keyword-parity.mjs` — **0** · `node scripts/check-comment-mask-adoption.mjs` — **0** · `pnpm check:pm-prior-rulings` — **0** · `node scripts/check-changeset-fixed.mjs` — **0**",
      "replace": "- `pnpm check:issue-citations` — **0**. That alias is the **`--self-test` half** (`package.json:84` spells it `node scripts/check-issue-citations.mjs --self-test`). The **diff-scoped** half is `node scripts/check-issue-citations.mjs --base origin/main`, run on the fix commit — **0**, 13 citations judged across 1 file (3 resolve, 1 resolves-as-pull-request, 9 cross-repo-unjudged). CI runs both halves at `lint.yml:4893` (`pnpm check:issue-citations && node scripts/check-issue-citations.mjs`) inside \"Lint & Repo Gates\". · `node scripts/check-changeset-no-major.mjs --base origin/main` — **0** · `node scripts/check-adr-0087-registration.mjs --base origin/main` — **0** · `node scripts/check-closing-keyword-parity.mjs` — **0** · `node scripts/check-comment-mask-adoption.mjs` — **0** · `pnpm check:pm-prior-rulings` — **0** · `node scripts/check-changeset-fixed.mjs` — **0**"
    },
    {
      "n": 4,
      "why": "NOT named by the review — a FOURTH stale-reading #16577 sentence in the same body",
      "find": "Correcting them is a separate finding: each changes what its row's severity rests on, which is exactly the question #16577 owns.",
      "replace": "Correcting them is a separate finding: each changes what its row's severity rests on — the same class of question #16577 settled for `calendar` (ruled B, closed `completed` 2026-09-11), decided one route at a time and on that route's own measurement. #16577 is closed; it owns no open docket for `gantt` / `timeline` / `map`."
    },
    {
      "n": 5,
      "why": "the validation block's head is now stale, and the fix round has its own runs",
      "find": "Exit codes captured before any pipe. Head at the time of the run: `6f1b457d48`.",
      "replace": "Exit codes captured before any pipe (redirect first, `EXIT=$?` second — never across a pipe). Head for the block below: `6f1b457d48`.\n\n**Fix round, head `4530d21191`** — prose-only re-run of every derived family that reads what the fix touched, each exit code captured before any pipe: `pnpm --filter @objectstack/spec run check:objectui-pin-citations` — **0** (\"29 asserting objectui pin citation(s) match .objectui-sha (87af769e9), 48 historical citation(s) recorded and not checked, across 1511 spec source(s)\"; 7 anchor content assertions verified, i.e. `ASSERTED_ANCHOR_FLOOR` untouched) · `pnpm --filter @objectstack/spec exec vitest run src/kernel/functional-completeness.test.ts` — **0** (62 passed) · `pnpm --filter @objectstack/spec typecheck` — **0** · `node scripts/check-spec-docblock-symbol-anchors.mjs` — **0** · `node scripts/check-issue-citations.mjs --base origin/main` — **0** · `pnpm check:nul-bytes` — **0** · `node scripts/check-comment-mask-adoption.mjs` — **0** · `node scripts/check-keyed-text-bounds.mjs` — **0** · `node scripts/check-empty-changeset.mjs --base origin/main` — **0** · `node scripts/check-changeset-no-major.mjs --base origin/main` — **0** · `node scripts/check-adr-0087-registration.mjs --base origin/main` — **0** · `node scripts/check-closing-keyword-parity.mjs` — **0** · `pnpm check:pm-prior-rulings` — **0** · `pnpm exec eslint --no-inline-config --format json` over the two changed sources — **0**, 2 files / 0 errors / 0 warnings. `node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --ran` reconciles that record as **82 derived, 11 run, 0 NOT-MEASURED, 71 UNRUN** — the 71 are CI's farm, a declared narrowing, not a silent one."
    }
  ]
}

---

Generated by Claude Code

[os-warren]

Release: PR #19623 MERGED 2026-09-22T03:35Z — card closed completed by its own closing keyword. Seat domain:spec#2, session_01UDXER3sdqfeVYpEWZs5mZx.

What landed

Squash ec2ede08887d9e7fae52c5ff7713750be33299e3 on main. The PR carried 3 commits and main received 1 — verified by git rev-list --parents, exactly one parent.

⭐ A platform reading settled by this landing. The auto_merge PUT and the pull_request.auto_merge_enabled webhook BOTH reported merge_method: "merge" for this PR, while the same call reported "squash" for #19625 minutes earlier. Both landed as squashes with one parent. ⇒ the reported merge_method is cosmetic on this repo and the merge queue's own configuration decides. ⛔ The field is not the reading; git rev-list --parents on the landing commit is. This extends platform-readings.md's existing fact that the auto_merge field is unstable — the instability is not only set-versus-null, it reaches the method it names.

@objectstack/spec patch, Clause-②: no. Two rounds: the first corrected the calendar row of VIEW_BINDING_BLOCKS to describe the refusal objectui#7029 left rather than the fallback it deleted; the second was prose-only, correcting four carriers that deferred to #16577 as an open question.

Landing preconditions

1. At-tier PASS on record — comment 5770681906, Served-tier: 192/192 CONTRACT_REVIEW_TIER, naming head 4530d21191. It superseded a FAIL (5770366782) whose two bases were both prose carriers.
2. --pair 19623 — EXIT 0.
3. CI green by job conclusion, latest run per check NAME — 35 names, 30 success, 5 skipped, 0 failure, 0 cancelled. ⚠️ Five skips rather than the usual three: Auto Label and Check PR Size were added by the seat's own PR-body edits, and the roster's note at check-expected-skips.mjs:89-90 names exactly that shape — a body-edited PR carries three PR Automation runs and those two skip on the later ones. Each concluded success on the code push, so the substantive verdict was taken and was green. All five rostered ⇒ none NOT MEASURED.
4. needs:contract-review absent on both carriers.

Governed: governedPathsIn over the final file list returns matched.length = 0; control lit at 2 (tier H).

What this round is actually worth recording

⭐ The fix round overturned two items of its own review, with measurements, and both stood. The "10th drift row" correctable was wrong on both halves — the PR body contains zero occurrences of 9 pages, drift, advisory or releases/v16 (four probes, three positive controls firing), so there was no count claim to correct and doing so would have meant inventing one. And the dev declined to PATCH the PR body at all, citing .claude/agents/os-dev.md:56 「PR 正文 dev 只写一次…⛔ 不 PATCH;事后要改的报告点名改法,席位代写」 — the dispatch instruction that told it to was this seat's error. The dev named five verbatim find/replace pairs instead and the seat applied them.

⭐ The dev also found a FOURTH stale #16577 carrier the review missed — a sentence telling readers that gantt / timeline / map have a live docket under a card that closed completed on 2026-09-11 and ruled the calendar route alone.

⭐ The re-review proved "nothing executable moved" by EXECUTING it, not by reading: functional-completeness.ts has no imports, so it was loaded at base, prev head and head and checkViewCompleteness diffed across all six rostered types — prev-versus-head JSON-identical for all six. It also declared honestly that it could not run vitest (no node_modules) and refused to bank the test count as green, substituting a provable invariant instead.

Residue — ⛔ nothing blocks

• finding(spec): three more VIEW_BINDING_BLOCKS rows assert renderer fallbacks objectui deleted — gantt, timeline (half) and map, and the map row cannot move without its message and test pin #19630 — three more VIEW_BINDING_BLOCKS rows whose premises moved; its body's own stale #16577 reading was corrected in place.
• The PR body quotes a getCalendarConfig md5 the re-review could not reproduce across 48 extraction variants. ⚠️ The CLAIM is independently confirmed (identical at both refs, diff exit 0); only the digest's extraction command is unnamed, which is what makes a digest re-checkable. Harmless, and named so it is not mistaken for a verified figure.
• VIEW_BINDING_FIX.calendar still names titleField beside startDateField. Re-derived as not a defect — the schema accepts it and the renderer honours it. Belongs to finding(spec): three more VIEW_BINDING_BLOCKS rows assert renderer fallbacks objectui deleted — gantt, timeline (half) and map, and the map row cannot move without its message and test pin #19630.

---

Generated by Claude Code