diff --git a/‎.changeset/9256-public-blocks-content-channels.md‎
Lines changed: 56 additions & 0 deletions b/‎.changeset/9256-public-blocks-content-channels.md‎
Lines changed: 56 additions & 0 deletions
diff --git a/‎packages/types/src/__tests__/content-channel-family-d-9256.test.ts‎
Lines changed: 27 additions & 12 deletions b/‎packages/types/src/__tests__/content-channel-family-d-9256.test.ts‎
Lines changed: 27 additions & 12 deletions
@@ -0,0 +1,56 @@
+---
+'@object-ui/types': minor
+---
+
+**BREAKING (shipped as `minor` — see below):** nineteen ADR-0080 public-block
+arms and the dashboard widget slot's `metric-card` node now refuse an authored
+`children` by name. None of these renderers reads the node's child list, so the
+key rendered nothing, with no render-time error or warning and no element
+(objectui#9256).
+
+- `page:header`, `page:tabs`, `page:accordion`, `record:details`,
+  `record:highlights`, `record:related_list`, `record:path`, `record:activity`,
+  `record:discussion`, `record:history`, `record:quick_actions`,
+  `record:reference_rail`, `element:text`, `element:number`, `element:button`,
+  `element:divider`, `object-metric` and `object-master-detail-form`: `children`
+  and `body` are declared as by-name refusals on the zod arm, each kept a member.
+  This package has no TypeScript declaration of these nodes, so the zod face is
+  the only one that changes.
+- `record:alert`: `children` only. Its `body` is the alert's message text, not a
+  content channel, and is left as it was.
+- `page:tabs` and `page:accordion`: the node's own `children` only. Each item's
+  `children` in `items` stays live — that is what these blocks render.
+- `metric-card` in a dashboard's `widgets`: `children?: never` and `body?: never`
+  on the TypeScript face (`DashboardWidgetSlotComponentSchema`), and both keys
+  refused by name on its zod twin. That twin is the first arm of the widget
+  slot's union, so the refusal reaches the author inside the union's
+  `invalid_union` issue at the widget's path, beside the strict widget schema's
+  `unrecognized_keys`; `objectui validate` prints it as one arm of two.
+
+What moves for an author:
+
+- `children` on any of these nodes parsed green (and, on `metric-card`, also
+  type-checked). It is now refused at `safeParse` time at its own path — on
+  `metric-card`, under the widget's `invalid_union` — and on
+  `DashboardWidgetSlotComponentSchema` at authoring time.
+- `body` was already refused on all of them, by `BaseSchema`. Everywhere except
+  `record:alert`, the zod refusal message now names what the node renders
+  instead, where it used to point at `children`, which these nodes do not read
+  either.
+
+No render behaviour changes: nothing read these keys, which is the whole reason
+they could be refused.
+
+Migration: each of these nodes renders from its own keys, so there is no channel
+to move the content to. Put it in the key the node does render
+(`element:text`'s `properties.content`, the item-level `children` of a
+`page:tabs` or `page:accordion` item, `record:alert`'s `properties.body`), place
+it beside the node in a container that reads `children` (`page:section`,
+`page:card`), or drop it.
+
+Also, text only: `CodeEditorSchema.children`'s docblock now names `onChange`
+among the keys `CodeEditorRenderer` forwards to Monaco.
+
+`minor` rather than `major` because this repo's version policy forbids `major`
+in any changeset — one `fixed` group — and records `minor` plus an explicit
+breaking note as the spelling for a breaking change here.
@@ -116,6 +116,13 @@
  *     RE-POINTED at a twin rather than inverted; what this card asserts — that it
  *     narrowed `children` on the chatbot faces and left their `body` alone — is
  *     unmoved.
+ *     ⚠️ AMENDED AGAIN (objectui#9256, public-block slice): the hold-out on the
+ *     two TWIN faces has ended too, and not by a decision on this card —
+ *     objectui#6771 retired `body` on `BaseSchema`, so each twin now declares
+ *     the same neither-channel tombstone on `body` that its `children` carries,
+ *     pointing at `requestBody`. `body` is REFUSED on all three chatbot faces
+ *     today; the two controls below assert that refusal (the mirror control)
+ *     and its compile-time twin, and neither proves `body` parses anywhere.
  *
  * ## ⚠️ Half of this file is a COMPILE-TIME assertion and vitest CANNOT read it
  *
@@ -301,17 +308,22 @@ const ROWS: ReadonlyArray<readonly [
   ['calendar-view', CalendarViewMirror as unknown as Mirror, ['body', 'children'], {}],
   // ⚠️ ONE-SIDED ROW, and ⛔ not a two-sided reading (objectui#9659, carrying a
   // contract-review residual on objectui#9639). The three rows below list `children`
-  // only, and on the two TWIN faces that still means what it always meant: `children`
-  // dead, `body` held out and LIVE, with the same-face LIVE CONTROL below proving it.
-  // On the PLAIN `chatbot` face it no longer does. Ruling A on objectui#8572 retired
-  // `ChatbotSchema.body` as an ADR-0049 tombstone for a DIFFERENT reason than this card's
-  // — a naming collision, not a dead content channel — and objectui#9639 landed it, so
-  // the plain face refuses BOTH channels today. Measured: `body` is ACCEPTED on
-  // `chatbot-enhanced` and `chatbot-floating`, REFUSED on `chatbot`.
-  // ⇒ the absence of `body` from the plain row means "not this card's to assert", ⛔ not
-  // "still live here", and objectui#9639 had to re-point both controls below at a twin
-  // precisely because no same-face control is available any more. The `body` half of the
-  // plain face is pinned by `node-recursion-point-8344.test.ts`, which owns objectui#8572.
+  // only. That no longer means `body` is live on any of them: `body` is REFUSED on all
+  // three faces today, for two different reasons.
+  //   - PLAIN `chatbot`: ruling A on objectui#8572 retired `ChatbotSchema.body` as an
+  //     ADR-0049 tombstone for a DIFFERENT reason than this card's — a naming collision,
+  //     not a dead content channel — and objectui#9639 landed it.
+  //   - The TWINS `chatbot-enhanced` / `chatbot-floating`: objectui#6771 retired `body`
+  //     on `BaseSchema`, so each twin declares the same neither-channel tombstone on
+  //     `body` that its `children` carries, pointing at `requestBody`. The LIVE CONTROL
+  //     below asserts that refusal on `chatbot-enhanced`.
+  // Re-measured in the objectui#9256 public-block slice: `body` is REFUSED on
+  // `chatbot`, `chatbot-enhanced` and `chatbot-floating` alike. (This comment used to
+  // say `body` was ACCEPTED on the two twins; that stopped being true when the twins
+  // gained their `body` tombstone, and the control below already asserted the refusal.)
+  // ⇒ the absence of `body` from these rows means "not this card's row to assert", ⛔ not
+  // "still live here". The `body` half of the plain face is pinned by
+  // `node-recursion-point-8344.test.ts`, which owns objectui#8572.
   // ⛔ Do not add `'body'` to the plain row to "fix" this: the message that row's
   // assertions read is objectui#9256's, and the plain face's tombstone carries
   // objectui#8572's instead — the row would go red on a true statement.
@@ -405,7 +417,10 @@ describe('objectui#9256 — CONTROLS: the node itself, and the held-out channel,
     expect(Object.keys(mirror.shape)).toContain(key);
   });
 
-  it('LIVE CONTROL — the chatbot family still accepts `body`, the channel held out of this card', () => {
+  // Named for what it asserts since the objectui#9256 public-block slice: the
+  // name used to say the family "still accepts `body`", which the body below has
+  // not asserted since the twins gained their `body` tombstone.
+  it('LIVE CONTROL — the chatbot twins REFUSE `body` too, pointing the author at `requestBody`', () => {
     // The held-out channel, still held out — RE-POINTED, not inverted, by
     // objectui#8572. This control was aimed at `ChatbotSchema`, whose own `body`
     // the parity ledger recorded as "two different meanings of one key — a naming