docs: cite landing commits where pending changesets and source comments name 23 objectui issues that answer 404, and re-qualify 19 bare objectstack numbers (objectui#10803, batch 6)

.changeset/10117-page-header-title-lookup-fallback.md

@@ -2,7 +2,7 @@
 '@object-ui/components': patch
 ---
 
-fix(components): `page:header` resolves a lookup title candidate instead of handing its expanded object to JSX (objectui#10117)
+fix(components): `page:header` resolves a lookup title candidate instead of handing its expanded object to JSX
 
 A record whose declared name field was empty took the whole `page:header` block
 down with "Objects are not valid as a React child" (React #31), leaving the red

.changeset/10120-form-omits-fls-denied-fields.md

@@ -3,7 +3,7 @@
 '@object-ui/fields': patch
 ---
 
-A form no longer submits — nor offers — a field the CALLER may read but not edit (objectui#10120).
+A form no longer submits — nor offers — a field the CALLER may read but not edit.
 
 **Clause-②: no** — no exported symbol is added, removed, renamed or retyped, no key on a published payload moves, and no accept set is relaxed. `sanitizeFormData` gains an optional third argument and `fieldWriteGate` / `applyFieldPermissions` are new, but `@object-ui/plugin-form` publishes `.` only, from `index.tsx`, which re-exports neither module — measured on the built `dist/index.d.ts`, where neither name appears. `LookupField`'s props are unchanged. What moves is what the client PUTS on the wire and which controls it draws.
 

.changeset/10129-action-param-field-backed-picker.md

@@ -5,8 +5,7 @@
 ---
 
 A field-backed action param reaches its record picker, and a param whose backing
-field cannot be read is refused instead of rendered as an empty text box
-(objectui#10129).
+field cannot be read is refused instead of rendered as an empty text box.
 
 An action declaring a `params` entry backed by a `lookup` field rendered as a bare
 text input: no options, no typeahead, and no request for the referenced object on

.changeset/10132-dashboard-i18n-declared-not-resolved.md

@@ -3,7 +3,7 @@
 '@object-ui/plugin-dashboard': patch
 ---
 
-fix(dashboard,charts): two dashboard surfaces the spec types as translatable now resolve (objectui#10132)
+fix(dashboard,charts): two dashboard surfaces the spec types as translatable now resolve
 
 `@objectstack/spec` declares both of these translatable and the Console resolved neither, so
 an author could set a key the contract documents and nothing happened. They turned out to be

.changeset/10163-line-items-field-write-gate.md

@@ -4,7 +4,7 @@
 
 A `record:line_items` grid no longer offers a cell the CALLER may read but not edit (objectui#10163).
 
-**What it was.** objectui#10120 taught the record-form containers to render a field the caller's permission set marks `editable: false` as disabled, through one shared render pass. The line-items panel stayed outside that pass: it read `readonly` and nothing else, so on the same page the same column rendered as a live, editable cell — inviting an edit the server refuses.
+**What it was.** `80c54122e` taught the record-form containers to render a field the caller's permission set marks `editable: false` as disabled, through one shared render pass. The line-items panel stayed outside that pass: it read `readonly` and nothing else, so on the same page the same column rendered as a live, editable cell — inviting an edit the server refuses.
 
 **What changed, in observable terms.**
 

.changeset/10563-form-arms-write-payload.md

@@ -7,7 +7,7 @@ The `tabbed`, `wizard` and `split` form layouts now write what the simple form w
 
 **Clause-②: no.** No exported symbol, type or prop changes. The shared sequence lives in an internal module that `index.tsx` does not re-export. The change is in what the client sends, and ⚠️ in what a host `submitHandler` receives from these layouts (see below).
 
-**Before.** The simple form, `ModalForm` and `DrawerForm` strip a save before sending it, and on an edit they send only the fields that differ from the record they read (objectui#10108, objectui#10120, objectui#10156). `TabbedForm`, `SplitForm` and `WizardForm` did neither. An edit sent every value the form held: `id`, `owner_id`, `created_by`, `updated_at`, formula columns, and fields the caller's field-level security refuses. The server answered with a `403` or an unknown-field refusal. A create seeded with a whole record, such as a copy of an existing one, posted those columns too. A simple form whose mobile `stepper` option shows it one step at a time renders through `WizardForm`, so it had the same defect.
+**Before.** The simple form, `ModalForm` and `DrawerForm` strip a save before sending it, and on an edit they send only the fields that differ from the record they read (objectui#10108, `80c54122e`, objectui#10156). `TabbedForm`, `SplitForm` and `WizardForm` did neither. An edit sent every value the form held: `id`, `owner_id`, `created_by`, `updated_at`, formula columns, and fields the caller's field-level security refuses. The server answered with a `403` or an unknown-field refusal. A create seeded with a whole record, such as a copy of an existing one, posted those columns too. A simple form whose mobile `stepper` option shows it one step at a time renders through `WizardForm`, so it had the same defect.
 
 **What changed, in observable terms.**
 

.changeset/10746-joined-report-clears-binding.md

@@ -9,8 +9,8 @@ to `joined`, and in the same render hid its dataset / values / rows / columns /
 controls, while the spec's own `reportForm` hid the whole "Dataset binding" section
 (`order` included) through `visibleWhen: "data.type != 'joined'"`. A report that was
 bound first and switched second kept every one of those keys invisibly. `ReportSchema`'s
-joined arm refuses a container `dataset` / `rows` / `columns` / `values` (objectstack PR
-#20160: "a `joined` report selects per block — move `KEY` onto `blocks[]`, or delete it")
+joined arm refuses a container `dataset` / `rows` / `columns` / `values` (PR
+objectstack#20160: "a `joined` report selects per block — move `KEY` onto `blocks[]`, or delete it")
 and has always refused a container `order`, so the save was refused at a path no control
 on the Properties tab could reach; only the JSON source tab could delete the key.
 

.changeset/10803-dead-citation-sweep-sixth-batch.md

@@ -0,0 +1,24 @@
+---
+---
+
+Comment-only in `@object-ui/app-shell`, `@object-ui/components`, `@object-ui/core`,
+`@object-ui/data-objectstack`, `@object-ui/fields`, `@object-ui/layout`,
+`@object-ui/plugin-dashboard`, `@object-ui/plugin-designer`, `@object-ui/plugin-detail`,
+`@object-ui/plugin-form`, `@object-ui/plugin-grid`, `@object-ui/plugin-kanban`,
+`@object-ui/plugin-list`, `@object-ui/plugin-view` and `@object-ui/types` (whose runtime
+text is declared separately below): docblocks and code comments that cited objectui issues
+which answer 404 now cite the commit that landed each change, as a 9-character sha, and
+where nothing answers they name the card by role (objectui#10803, the sixth batch). Bare
+numbers that were sister-repo cards or pull requests are now written `objectstack#N`, and
+where the sister-repo pull request itself answers 404 the sentence cites its commit in
+that repository instead. Two `@object-ui/plugin-grid` sites beside `objectstack#3720` that
+named the apiMethods whitelist card bare now read `objectstack#3391`, the card they meant.
+Some of these docblocks reach the
+emitted `.d.ts` and `.js`; none of them moves a claim, and none of these comment edits
+changes a code or type token. No published behaviour changes through them, so this
+declares no release. The one runtime string the sweep touches, a zod `.describe()` string
+in `@object-ui/types`, is declared on its own as a patch in
+`10803-sixth-batch-runtime-strings.md`.
+
+The same repair in the pending changesets that carried these citations is prose-only, and
+their frontmatter is byte-identical.

.changeset/10803-sixth-batch-runtime-strings.md

@@ -0,0 +1,14 @@
+---
+'@object-ui/types': patch
+---
+
+fix(types): `InputSchema.wrapperClass`'s description no longer points at an objectui issue that answers 404
+
+The zod `.describe()` text of `InputSchema.wrapperClass` ended with a pointer to an
+objectui issue that answers 404. A reader of a zod `description` has no repository to
+resolve a commit against, so the pointer is dropped rather than replaced: the description
+now reads "Classes on the wrapper div around the input and its label" (objectui#10803).
+
+Nothing else in the string moves, and no key, path, issue code, accept set, refusal or
+severity changes: every document that parsed before parses the same way and is refused at
+the same path with the same code.

.changeset/14026-list-import-mappings-pin.md

@@ -2,5 +2,5 @@
 ---
 
 Pin `listImportMappings(objectName)` against the framework's measured
-`GET /api/v1/meta/mapping` body (objectui#14026). Test only; no package is
+`GET /api/v1/meta/mapping` body (objectstack#14026). Test only; no package is
 released by this change.

.changeset/6594-headercolor-mirror-enum.md

@@ -8,7 +8,7 @@ the contract — the TypeScript declaration and the `@object-ui/types/zod` mirro
 objectstack#12126). The six are `muted`, `muted/50`, `accent`, `primary/10`, `secondary/10`
 and `destructive/10`: exactly what `@object-ui/plugin-detail`'s `HEADER_COLOR_CLASSES`
 resolves (objectui#6178) and exactly what `@objectstack/spec` declares on its strict
-`record:details` section schema (objectstack PR #12616).
+`record:details` section schema (PR objectstack#12616).
 
 ## ⚠️ Accept-set narrowing — these spellings stop validating
 

.changeset/6748-preview-mode-provenance-ratchet.md

@@ -5,7 +5,7 @@ Test-only change in `@object-ui/auth`: `auth-spec-parity.test.ts` now watches
 `PreviewModeConfig`, the `@objectstack/spec/kernel` symbol that
 `packages/auth/README.md:328` claims the preview-mode prop aligns with. Upstream
 retired that symbol in source and registered it as `kernel/PreviewModeConfig` in
-`RETIRED_DEFS_BY_MAJOR[18]` (objectstack#11846, landed as objectstack PR #12718), so
+`RETIRED_DEFS_BY_MAJOR[18]` (objectstack#11846, landed as PR objectstack#12718), so
 it leaves the published set at spec major 18 — it is still exported by the 17.2.0 this
 repo resolves, which is why the assertion ships green. It goes red at the spec-18 bump,
 the exact moment the README sentence stops being true, which is the signal to correct

.changeset/6985-wizard-card-r-alignment.md

@@ -6,7 +6,7 @@
 
 Wizard view v1, the objectui half (Card R, objectui#6985) — alignment + pins for the
 ruled `type: 'wizard'` tightening (objectstack#13622 D1–D8, maintainer ruling
-2026-08-31; spec half objectstack PR #13733).
+2026-08-31; spec half PR objectstack#13733).
 
 The renderer was already aligned: `WizardStepConfig` carries no predicate/collapse
 keys (objectui#6237's ruled split), the wizard route drops-and-reports an authored
@@ -42,5 +42,5 @@ exemption since #2959. This card lands the residue:
   half records the 17.2.x accept-set it measured. `steps:` is pinned refused on
   every spec line.
 
-No teaching material — the #13337/#13086 fence lifts only after both halves land;
+No teaching material — the objectstack#13337 / objectstack#13267 fence lifts only after both halves land;
 docs changes here are TSDoc/comments only.

.changeset/7415-action-type-input-rename.md

@@ -19,7 +19,7 @@ actually set it:
   last, so `type="api"` replaced the component discriminator and the node stopped resolving
   to a component at all. `validate.ts` cannot report that: `type` is in `BASE_PROPS`, so it
   is skipped before the declared-input check runs. Two mechanisms, one outcome, no
-  diagnostic. objectstack PR #14274 landed a refusal on this tier whose prescription
+  diagnostic. PR objectstack#14274 landed a refusal on this tier whose prescription
   ("write the tag you meant") is wrong for exactly these two components.
 - **react-page tier** — the wrapper stamps `type: tag` last and parks the author's value
   under `specType` (objectui#2880), which neither action renderer reads.

.changeset/7714-lookup-draft-stays-client-side.md

@@ -31,7 +31,7 @@ objectstack#4001 closed.
 
 **The whitespace row follows the contract; it is not a local opinion.** The predicate
 is `typeof reference === 'string' && reference.trim() !== ''`. It was a declared
-divergence when written — 17.3.0's #13632 refinement spelled its emptiness test as an
+divergence when written — 17.3.0's objectstack#13632 refinement spelled its emptiness test as an
 equality against `''`, so the spec accepted `reference: '   '` while these writers
 refused it. objectstack#16920 applies that test to the TRIMMED value, so the spec now
 refuses the identical shape under the same `custom` issue at the same `reference`

.changeset/7722-wrapper-class-five-more.md

@@ -43,6 +43,6 @@ admitted unexamined on all five mirrors, pinned per mirror with a control key
 the renderer does not read. `InputSchema.wrapperClass`, declared on the TS face
 only, was a recorded row of the parity ledger (`UnmirroredDeclared`) and this
 card left it there; the new sweep pin carried it as a self-expiring exemption.
-objectui#8072 has since mirrored that key, so the row and the exemption are both
+`c974edf14` has since mirrored that key, so the row and the exemption are both
 gone — the exemption expired exactly as designed, and the sweep now judges all
 nine readers alike.

.changeset/7735-zod-mirrors-stop-authoring-defaults.md

@@ -26,7 +26,7 @@ was not even self-consistent.
 **What changed.** All 41 `.default()` call sites under `packages/types/src/zod/`
 are removed — `layout.zod.ts` 22, `crud.zod.ts` 11, `form.zod.ts` 5,
 `views.zod.ts` 2, `app.zod.ts` 1. `@object-ui/components` reconciles the third
-face objectui#8229 found: `flex`'s registration `defaultProps.align` seeded
+face a separate finding found: `flex`'s registration `defaultProps.align` seeded
 `'center'`, the value its own renderer never applies, so a designer-made node
 laid out differently from a hand-authored one; it now seeds `'start'`.
 

.changeset/7741-list-import-mappings-discriminate.md

@@ -40,6 +40,6 @@ a channel added ALONGSIDE that contract, not a change to it.
   user without devtools open can tell "there are none" from "we could not find out".
   Three new `console.importMappings*` keys ship in all ten locale packs.
 
-This applies framework #13906 decision 1 option A — *a thing that could not be READ is
+This applies framework objectstack#13906 decision 1 option A — *a thing that could not be READ is
 not a thing that is ABSENT* — at this seam. It is an already-adopted discrimination, not
 a new principle.

.changeset/8072-input-wrapper-class-mirrored.md

@@ -2,7 +2,7 @@
 '@object-ui/types': patch
 ---
 
-Mirror `wrapperClass` on `InputSchema` (objectui#8072 — the last
+Mirror `wrapperClass` on `InputSchema` (the last
 `schema.wrapperClass` reader whose value the validator admitted unexamined).
 
 `packages/components/src/renderers/form/input.tsx` reads

.changeset/8137-record-activity-calibration-repoint.md

@@ -2,7 +2,7 @@
 ---
 
 Re-point the `record:activity.types` member calibration control in
-`registry-inputs-spec-parity` from `'Account'` to `''` (objectui#8137).
+`registry-inputs-spec-parity` from `'Account'` to `''`.
 `@objectstack/spec` 17.3.0 made that vocabulary open —
 `z.array(z.union([FeedItemType, z.string().min(1)]))`, objectstack#11658
 executing the maintainer's 2026-08-24 ruling on objectstack#11507 — so the old

.changeset/8204-refuseskind-invalid-type-decisive.md

@@ -5,7 +5,7 @@ Test-only change: repair the `registry-inputs-spec-parity` kind classifier so a
 whose contract is a `z.union` is judged by what the union actually refused. No published
 behaviour changes; nothing outside `apps/console/src/__tests__/` is touched.
 
-objectui#8204. `refusesKind` asked "is EVERY complaint at this node a kind complaint",
+`refusesKind` asked "is EVERY complaint at this node a kind complaint",
 and Zod keeps running a schema's checks after its type check has failed: for
 `z.string().min(1)` the probe `[]` comes back as both `invalid_type` AND `too_small`.
 Read by `every`, that pair is a CONTENT refusal — so the array probe that

.changeset/8236-collapsible-open-intercept-retire.md

@@ -26,7 +26,7 @@ change the block runs on its own published `inputs` (`defaultOpen` / `disabled`
 
 **The order is load-bearing and is not an implementation detail.** The renderer-side
 exclusion lands FIRST and the retirement second, because the render path runs no
-`safeParse` (objectui#9585 measured it NOT GATED). Retiring the declaration alone would
+`safeParse` (a pin in this change measures it NOT GATED). Retiring the declaration alone would
 have deleted the author's only warning while leaving the takeover running — strictly
 worse than doing nothing. Both halves are in this change; ⛔ neither is safe to remove
 alone.

.changeset/8291-workspace-admin-reads-the-rung.md

@@ -29,7 +29,7 @@ the two scalar legs, where they are the raw membership role and the stored `user
 
 The docblock that argued AGAINST reading `user.isPlatformAdmin` — on the grounds that the server
 computed it as `'platform_admin' in positions`, "two spellings for one fact" — is rewritten rather
-than left standing. After #15948 the flag derives from the unscoped `admin_full_access` grant and
+than left standing. After objectstack#15948 the flag derives from the unscoped `admin_full_access` grant and
 the array does not, so they can disagree and the rung is right.
 
 No behaviour changes for a genuine administrator: a platform administrator on a single-tenant

.changeset/8307-kanban-lane-count-honesty.md

@@ -2,8 +2,7 @@
 '@object-ui/plugin-kanban': patch
 ---
 
-A Kanban lane header over a windowed fetch now says `77+` instead of `77`
-(objectui#8307).
+A Kanban lane header over a windowed fetch now says `77+` instead of `77`.
 
 **The defect.** `object-kanban` fetches with a real `$top` (objectui#4025) and then
 groups **what came back** into lanes client-side, so `col.cards.length` counts fetched

.changeset/8408-public-form-title-description.md

@@ -3,7 +3,7 @@
 ---
 
 Public form `/f/:slug` renders its authored `title` and `description` instead of the
-object API name (objectui#8408).
+object API name.
 
 The one Console surface an **unauthenticated** visitor sees greeted them with a
 database table name. `GET /api/v1/forms/:slug` serves the authored copy intact

.changeset/8458-parity-header-pairs-figure.md

@@ -9,7 +9,7 @@ so the file never reaches a consumer. Declared as releasing nothing.
 - objectui#8458: the WIDER ledger's objectui#7760 movement sentence said "16 pairs",
   the one figure in that paragraph no `headerFigures` pin reached. Re-derived as 18
   (13 emptied + 5 reduced), defined, and pinned.
-- objectui#8248: objectui#8222's exclusion note called the seed decomposition a LIVE
+- objectui#8222's exclusion note called the seed decomposition a LIVE
   figure, which objectui#8243 made historical. The guidance clause is amended per the
   maintainer ruling (decision batch #71); the two measurement-record clauses are
   untouched, and the record-vs-guidance rule is now written once in the file header.

.changeset/8603-record-details-hide-empty-restored.md

@@ -13,6 +13,6 @@
 
 **What did NOT change.** The empty ROWS of a section that still has a filled row stay with `DetailSection`'s auto-hide heuristic and the reader's "Show N empty fields" toggle, which no authored value overrides in either polarity (objectui#7129 Q2-C, left standing by the ruling). Inline-edit mode renders an all-empty section either way, so its fields never become unreachable. `record:reference_rail`'s own component-level `hideEmpty` and the `detail.hideEmptyFields` toggle label are different keys and are untouched.
 
-**Why it moved twice.** objectui#7129 (maintainer 2026-09-01) retired the declaration and the read because `@objectstack/spec` REFUSED the key — true at the 17.2.0 pin this repo held. Upstream had already declared it (17.3.0, upstream #11289, maintainer ruling 2026-08-23 direction 1, taken from a measured symptom), so once the pin moved the contract promised an author a behaviour the renderer no longer had, and it parsed green at publish. objectui#8603 (director seat batch #137 item 3, maintainer 2026-09-15) ruled the protocol correct and restored the read; #7129's Q1-A is superseded for this key only.
+**Why it moved twice.** objectui#7129 (maintainer 2026-09-01) retired the declaration and the read because `@objectstack/spec` REFUSED the key — true at the 17.2.0 pin this repo held. Upstream had already declared it (17.3.0, upstream objectstack#11289, maintainer ruling 2026-08-23 direction 1, taken from a measured symptom), so once the pin moved the contract promised an author a behaviour the renderer no longer had, and it parsed green at publish. objectui#8603 (director seat batch #137 item 3, maintainer 2026-09-15) ruled the protocol correct and restored the read; #7129's Q1-A is superseded for this key only.
 
 All four parties now agree — the spec declares it, `@object-ui/types` declares it, the `DetailViewSectionSchema` zod mirror carries it, and the renderer reads it — pinned together in `packages/plugin-detail/src/renderers/__tests__/record-details.hideEmptyRetired-7129.test.tsx`.