objectstack-ai
diff --git a/‎.changeset/field-value-domain-write-path.md‎
Lines changed: 65 additions & 0 deletions b/‎.changeset/field-value-domain-write-path.md‎
Lines changed: 65 additions & 0 deletions
diff --git a/‎content/docs/data-modeling/validation-rules.mdx‎
Lines changed: 1 addition & 1 deletion b/‎content/docs/data-modeling/validation-rules.mdx‎
Lines changed: 1 addition & 1 deletion
diff --git a/‎packages/lint/src/validate-predicate-path-refs.test.ts‎
Lines changed: 23 additions & 8 deletions b/‎packages/lint/src/validate-predicate-path-refs.test.ts‎
Lines changed: 23 additions & 8 deletions
diff --git a/‎packages/objectql/src/validation/record-validator.ts‎
Lines changed: 59 additions & 0 deletions b/‎packages/objectql/src/validation/record-validator.ts‎
Lines changed: 59 additions & 0 deletions
@@ -0,0 +1,65 @@
+---
+'@objectstack/objectql': minor
+'@objectstack/spec': minor
+---
+
+feat(objectql,spec): `Field.valueDomain` binds at the write seam — a non-member is refused with `value_domain` (maintainer ruling 2026-09-02 on #14168, engine half)
+
+**BREAKING** accept-set narrowing on the ObjectQL record write path, shipped as
+`minor` under the repo's launch-window convention for breaking changes.
+
+The key is **already published, and published unenforced**. The version-packages
+cut `8a1bad8b8` (2026-09-04 10:20Z) consumed the spec half's changeset
+`field-value-domain-slot.md` and released `@objectstack/spec@17.3.0`, which
+declares `Field.valueDomain`, parses it, and refuses it on any type other than
+`text` — and never reads it when a record is written. The 17.3.0 liveness ledger
+states the gap in its own words: "a non-member WRITTEN to a `text` field
+declaring a domain is accepted today". That write is accepted on 17.3.0 and is
+refused from this release on.
+
+**Refused shape**, precisely: a record write that supplies a value for a `text`
+field whose definition declares `valueDomain`, where the WRITTEN value is not a
+member of the named standard. It fails with the field error code `value_domain`,
+carrying `constraint: { valueDomain }` and a message that names the standard in
+all four platform locales. Nothing else narrows — a field that declares no
+`valueDomain` is untouched, and so is every other field type, because the schema
+accepts the key on `text` alone and the validator judges exactly that set.
+
+**Remedy: write a member of the declared standard.** `iana_time_zone` admits
+`UTC` and refuses `Mars/Olympus`; `iso_4217_currency` admits `CHF` and refuses
+`chf`; `iso_3166_alpha2` admits `CH` and refuses `ZZ`. Dropping the
+`valueDomain` declaration from the field lifts the refusal entirely, for an
+author who declared a domain they did not mean.
+
+**No stored row is touched, and none becomes invalid.** This is the `min` /
+`max` / `maxLength` transition-gate class: a value stored before the domain was
+declared — or before this release — is never re-read, and it survives an edit of
+another field on the same record. An absent or empty value follows the field's
+`required` handling, not this check.
+
+<!-- adr-0087: not-required (no-migration-prescription) Nothing authorable is
+renamed, retired or tombstoned. `Field.valueDomain` keeps its name, its type and
+its position; this release only makes the declaration the key already carries
+bind at the write seam, so `objectstack migrate meta` has no metadata to
+rewrite — a document that declares a domain is already in its final spelling,
+and one that declares none is untouched. ⚠️ This disposition does NOT rest on
+the key being unpublished, and must not be read that way: 17.3.0 shipped
+`Field.valueDomain` declared, parsed and UNENFORCED, which is exactly why this
+changeset carries the BREAKING banner above. It rests on the stored side
+instead. A stored value outside a declared domain is never re-read, so no stored
+row is invalidated here and none is reachable by a ledger entry at all. And
+which member a stored non-member SHOULD have been is authoring intent no ledger
+entry can decide: the stored string carries no evidence of whether the author
+meant a different member of that standard, a different standard, or no
+declaration at all. The channel that reaches the author is the refusal itself,
+raised at the write, naming the standard — the same ground the sibling
+accept-set narrowing #15319 stands its own `no-migration-prescription`
+disposition on. -->
+
+- The membership test is the spec's shared `isValueDomainMember` — the same
+  predicate, over the same closed vocabulary, that a settings specifier's
+  `valueDomain` uses. A time zone accepted in Settings is the time zone
+  accepted in a field.
+- The two authoring forms (`fieldForm`, `objectForm`) gain a `valueDomain`
+  control, shown on exactly the types the schema accepts the key on. The
+  object-form control's choices are derived from the vocabulary, not re-typed.
@@ -44,7 +44,7 @@ These properties apply to **all** field types and are validated by the base `Fie
 | `maxLength` | `number` | — | Rejects values exceeding character count |
 | `minLength` | `number` | — | Rejects values below character count |
 | `format` | `string` | — | Validates against format pattern (e.g., regex) |
-| `valueDomain` | `'iana_time_zone' \| 'iso_4217_currency' \| 'iso_3166_alpha2'` | — | Constrains the written value to a published standard — an IANA time zone (judged by the `Intl.DateTimeFormat` probe, so `UTC` and `Asia/Kolkata` are members and `Europe/Munich` is not), an ISO 4217 currency code or an ISO 3166-1 alpha-2 country code (both exact uppercase). Membership, not shape: a pattern such as `^[A-Z]{2}$` admits `ZZ`; the domain does not. The same closed vocabulary and the same membership test as a settings specifier's `valueDomain`; a non-member is refused on the write path with the field error code `value_domain` (the engine half of the same ruling — until it lands, the declaration is accepted at parse and not yet enforced on writes). `text` only — declaring it on any other type is refused at parse. |
+| `valueDomain` | `'iana_time_zone' \| 'iso_4217_currency' \| 'iso_3166_alpha2'` | — | Constrains the written value to a published standard — an IANA time zone (judged by the `Intl.DateTimeFormat` probe, so `UTC` and `Asia/Kolkata` are members and `Europe/Munich` is not), an ISO 4217 currency code or an ISO 3166-1 alpha-2 country code (both exact uppercase). Membership, not shape: a pattern such as `^[A-Z]{2}$` admits `ZZ`; the domain does not. The same closed vocabulary and the same membership test as a settings specifier's `valueDomain`; a non-member is refused on the write path with the field error code `value_domain`. `text` only — declaring it on any other type is refused at parse. |
 
 **Default constraints:** None. Unbounded text unless `maxLength` is set.
 
 
@@ -535,12 +535,22 @@ describe('#7010 corpus — shipped METADATA_FORM_REGISTRY', () => {
     // a `page` section to `view.form.ts` — the surface block for the new `page`
     // view type, gated by `visibleWhen: "data.type == 'page'"` exactly as every
     // other surface block is — so the walk has one more predicate to reach.
-    // It is 51 today: objectui#6140 (maintainer ruling 2026-08-25, Option A)
+    // It was 51 after objectui#6140 (maintainer ruling 2026-08-25, Option A)
     // declared `rows` on the multiline editor types, adding one
     // `data.type in […]`-gated row to the field form AND one to the object
     // form's fields repeater — two more predicates for the walk to reach.
-    // Earlier measurements stay what they were: history, not the census.
-    expect(predicates, 'the shipped metadata forms carry no predicates at all').toBe(51);
+    // It is 53 today, and for the same shape: the maintainer ruling 2026-09-02
+    // (option A on the field-level `valueDomain`) put a `valueDomain` row in
+    // both authoring forms, each gated `data.type in ['text']` — the applicable
+    // type set, mirrored from the schema's own `VALUE_DOMAIN_FIELD_TYPES`.
+    // Measured rather than inferred from the delta: the corpus was enumerated
+    // on this tree and on the merge base, and differenced by
+    // `<form>::<field>::<source>` rather than by array index (inserting a row
+    // shifts every later sibling's index, which is churn, not corpus change).
+    // Exactly two entries were added — `field :: valueDomain` and
+    // `object :: valueDomain`, both `data.type in ['text']` — and NONE was
+    // removed. Earlier measurements stay what they were: history, not the census.
+    expect(predicates, 'the shipped metadata forms carry no predicates at all').toBe(53);
 
     const findings = validatePredicatePathRefs(corrupted);
     expect(findings).toHaveLength(predicates);
@@ -640,9 +650,14 @@ describe('#7010 corpus — shipped METADATA_FORM_REGISTRY', () => {
     // no longer offered a `set_null` the schema refuses. It is 18 today:
     // objectui#6140 added a `rows` row to the object form's fields repeater,
     // gated by `data.type in ['textarea','markdown','html','richtext']` — one
-    // more `data.type`-rooted predicate for the debare walk to restore. #6254's
-    // own measurement was 16 and stays 16 — that number is history, this one
-    // is a census.
+    // more `data.type`-rooted predicate for the debare walk to restore. It is 19
+    // today: the maintainer ruling 2026-09-02 (option A on the field-level
+    // `valueDomain`) added a `valueDomain` row to the same repeater, gated
+    // `data.type in ['text']`. That row is the ONLY object-form predicate this
+    // tree adds over its merge base — the two corpora were enumerated and
+    // differenced by `<form>::<field>::<source>`, and the object-form half of
+    // the two-entry delta is exactly it. #6254's own measurement was 16 and
+    // stays 16 — that number is history, this one is a census.
     const objectForm = structuredClone(METADATA_FORM_REGISTRY.object) as Record<string, unknown>;
     let restored = 0;
     const debare = (node: unknown): void => {
@@ -672,10 +687,10 @@ describe('#7010 corpus — shipped METADATA_FORM_REGISTRY', () => {
     expect(
       restored,
       "the object form's `data.type`-rooted predicates are no longer where this test looks",
-    ).toBe(18);
+    ).toBe(19);
 
     const findings = validatePredicatePathRefs({ views: [objectForm] });
-    expect(findings).toHaveLength(18);
+    expect(findings).toHaveLength(19);
     expect(new Set(findings.map((f) => f.rule))).toEqual(new Set([PREDICATE_PATH_UNROOTED]));
     expect(findings[0].message).toContain('`type`');
   });
 
@@ -22,6 +22,9 @@
  *                   multi-value field `[]` is an empty value (#9476 — the
  *                   #9447 ruling: required means non-empty array).
  *  - `maxLength` / `minLength`            (text/textarea/email/url/phone/password)
+ *  - `valueDomain`  a declared standard domain's membership, judged by the
+ *                   spec's shared `isValueDomainMember` — the WRITTEN value
+ *                   only (#14168, maintainer ruling 2026-09-02 option A)
  *  - `min` / `max`                        (number/currency/percent/rating/slider)
  *  - `scale`        more decimal places than declared → `max_scale` (#7501;
  *                   rejection, NEVER rounding — maintainer ruling 2026-08-11)
@@ -48,11 +51,13 @@ import {
   ALL_OPERATORS,
   RETIRED_FILTER_OPERATORS,
   BOUNDED_STRING_FIELD_TYPES,
+  VALUE_DOMAIN_FIELD_TYPES,
   REFERENCE_VALUE_TYPES,
   FILE_REFERENCE_TYPES,
   STRUCTURED_JSON_TYPES,
 } from '@objectstack/spec/data';
 import type { FieldErrorCode } from '@objectstack/spec/api';
+import { isValueDomainMember, type ValueDomain } from '@objectstack/spec/shared';
 import {
   renderValidationMessage,
   objectFieldLabelKey,
@@ -168,6 +173,15 @@ interface FieldDef {
   max?: number;
   /** Max decimal places for number types — enforced by rejection (#7501). */
   scale?: number;
+  /**
+   * Standard value domain the WRITTEN value must be a member of (#14168) —
+   * the same closed vocabulary and the same membership predicate a settings
+   * specifier's `valueDomain` uses, so a time zone accepted in Settings is the
+   * time zone accepted in a field. Typed as the spec's `ValueDomain` rather
+   * than `string`: an unknown domain word has no membership test to run, and
+   * `isValueDomainMember` is a total function over exactly this union.
+   */
+  valueDomain?: ValueDomain;
   options?: Array<{ value: string | number; label?: string } | string | number>;
 }
 
@@ -573,6 +587,51 @@ function validateOne(
     if (def.minLength !== undefined && s.length < def.minLength) {
       return fail('min_length', { minLength: def.minLength, actual: s.length });
     }
+    // ── `valueDomain` — membership in a published standard (#14168) ──
+    // Maintainer ruling 2026-09-02 (option A): ONE closed vocabulary and ONE
+    // membership predicate, shared by settings specifiers and object fields —
+    // so a currency code accepted in Settings is the code accepted in a field.
+    // The predicate is imported, never re-implemented: the repo already carries
+    // hand-rolled copies of the IANA probe, and a second opinion on membership
+    // is how "accepted in Settings, refused in a field" happens.
+    //
+    // The applicability door is the SPEC'S `VALUE_DOMAIN_FIELD_TYPES`, read as
+    // a constant for the same reason this branch reads `BOUNDED_STRING_FIELD_TYPES`
+    // (#11875): two seams reading one constant cannot drift into two opinions.
+    // The set is a strict subset of the bounded-string family (`text` alone
+    // today, against twelve) — which is why the check sits inside this branch
+    // and why the subset relation is pinned as a test rather than assumed.
+    // `FieldSchema` refuses the key outside the set at parse with a located
+    // issue at [valueDomain], and its refusal message states this seam's half
+    // of the contract verbatim: "the write-time validator applies `valueDomain`
+    // to exactly those types". Judging a hand-built runtime schema's key on the
+    // other eleven would make that sentence false.
+    //
+    // WRITTEN VALUE ONLY — the `min`/`max`/`maxLength` transition-gate class: a
+    // stored value outside a domain declared later is never re-read and survives
+    // unrelated edits (an omitted field never reaches here on update), and an
+    // absent/empty value is the `required` check's business above, not this one.
+    if (
+      def.valueDomain !== undefined &&
+      VALUE_DOMAIN_FIELD_TYPES.has(t) &&
+      !isValueDomainMember(def.valueDomain, s)
+    ) {
+      // One wire code — the ADR-0114 catalog member `value_domain`, with the
+      // domain shipped in `constraint` so a client can name it. The finer
+      // per-domain message key spells the standard out for a human ("a valid
+      // ISO 4217 currency code, e.g. CHF") in all four locales; it is a
+      // RENDERING choice and never reaches the wire, the same code/messageKey
+      // split `invalid_value_shape` and `required_cleared` use. The value is
+      // echoed because every one of those templates interpolates `{{value}}` —
+      // an uninterpolated placeholder ships `{{value}}` to the user verbatim.
+      return fail(
+        'value_domain',
+        { valueDomain: def.valueDomain },
+        `value_domain_${def.valueDomain}`,
+        undefined,
+        s,
+      );
+    }
     if (t === 'email' && !EMAIL_RE.test(s)) {
       return fail('invalid_email');
     }