@@ -0,0 +1,12 @@
+---
+'@objectstack/plugin-approvals': patch
+---
+
+Clause-②: no
+
+fix(plugin-approvals): an approval request's record title reads the flow's label, not `flow:NAME`
+
+- **`sys_approval_request.display_title` reads `process_label`.** The title is now the flow's label and the record id (`Discount Sign-off · opp_45`), the same label the request lists show. A row with no label stored falls back to `process_name` (`flow:deal_desk · opp_42`), so the title never reads `null · …`. Every surface that draws the request's title through `nameField` shows it: the record page's heading and breadcrumb, and the request cell of the action history list.
+- **`titleFormat` still names `{process_name}`.** A placeholder template cannot fall back, and `process_label` can be null after the boot fill (a row it could not label, or a failed fill run). A renderer that reads `titleFormat` ahead of `nameField`, such as the global search palette, still shows the machine source.
+
+A patch: one formula on one platform object's computed title. No export, type, accepted value or stored column changes.
@@ -22,6 +22,14 @@
  *
  * Where a source column is nullable, a row without it is titled by the other
  * column rather than failing to evaluate.
+ *
+ * `sys_approval_request` is the one object whose formula and `titleFormat`
+ * name different columns (#24007): the formula's flow half is the flow's label
+ * (`process_label`, nullable), with `process_name` as its fallback, while
+ * `titleFormat` keeps `{process_name}` because a template cannot express that
+ * fallback. So for that object point 2 is asserted where they agree — a row
+ * with no stored label, where both draw `process_name` — and a row with a
+ * label pins the formula's own text and the template's difference from it.
  */
 
 import { describe, it, expect, beforeAll, afterAll } from 'vitest';
@@ -90,7 +98,26 @@ describe('[#20015] approval objects resolve a real record title under ADR-0079 o
     }
   });
 
-  it('sys_approval_request: the H1 is "{process_name} · {record_id}", not the id', async () => {
+  it('sys_approval_request: with a flow label stored, the H1 is "{process_label} · {record_id}"', async () => {
+    // An authored label that is not the humanized machine name ("Deal Desk"),
+    // so the title can only have come from the stored label.
+    const created = await engine.insert('sys_approval_request', {
+      process_name: 'flow:deal_desk', process_label: 'Discount Sign-off',
+      object_name: 'opportunity', record_id: 'opp_45', status: 'pending',
+    }, SYS);
+    expect(created.display_title).toBe('Discount Sign-off · opp_45');
+    const row = await engine.findOne('sys_approval_request', { where: { id: created.id } }, SYS);
+    expect(row).not.toBeNull();
+
+    expect(h1Of(SysApprovalRequest, row!)).toBe('Discount Sign-off · opp_45');
+    expect(h1Of(SysApprovalRequest, row!)).not.toBe(row!.id);
+    expect(resolveRecordTitle(SysApprovalRequest, storedOnly(row!))).toBe('Discount Sign-off · opp_45');
+    // The pairing, restated: `titleFormat` keeps the machine source, so a
+    // renderer that reads it first draws a different text for this row.
+    expect(renderTitleFormat(SysApprovalRequest.titleFormat, row!)).toBe('flow:deal_desk · opp_45');
+  });
+
+  it('sys_approval_request: with no flow label stored, the H1 is "{process_name} · {record_id}", never "null · …"', async () => {
     const created = await engine.insert('sys_approval_request', {
       process_name: 'flow:deal_desk', object_name: 'opportunity', record_id: 'opp_42',
       status: 'pending',
@@ -99,6 +126,7 @@ describe('[#20015] approval objects resolve a real record title under ADR-0079 o
     expect(created.display_title).toBe('flow:deal_desk · opp_42');
     const row = await engine.findOne('sys_approval_request', { where: { id: created.id } }, SYS);
     expect(row).not.toBeNull();
+    expect(row!.process_label ?? null).toBeNull();
 
     expect(h1Of(SysApprovalRequest, row!)).toBe('flow:deal_desk · opp_42');
     expect(h1Of(SysApprovalRequest, row!)).not.toBe(row!.id);
 
@@ -45,12 +45,19 @@ export const SysApprovalRequest = ObjectSchema.create({
   isSystem: true,
   managedBy: 'engine-owned',
   description: 'Live approval instance tracked per submission',
-  // [ADR-0079] The record title is `display_title`, a text formula over the
-  // same two columns `titleFormat` names. The pointer used to be `id`: once a
-  // renderer honours ADR-0079's order (an explicit `nameField` wins over
-  // `titleFormat`), that made the record page's H1 the raw id. `titleFormat`
-  // stays for renderers that still read it first;
-  // `sys-approval-display-title.test.ts` holds the two to the same text.
+  // [ADR-0079] The record title is `display_title`, a text formula. The
+  // pointer used to be `id`: once a renderer honours ADR-0079's order (an
+  // explicit `nameField` wins over `titleFormat`), that made the record page's
+  // H1 the raw id. The formula reads the flow's label (`process_label`) and
+  // falls back to the machine source (`process_name`) where no label is
+  // stored. `titleFormat` stays for renderers that still read it first — the
+  // global search palette (`searchAll` in `@objectstack/metadata-protocol`)
+  // ranks it above `nameField` — and keeps `{process_name}`: a placeholder
+  // template cannot say "else", and `process_label` can be null after the boot
+  // fill (a row the fill could not label, or a failed fill run), where a
+  // `{process_label}` template draws a title with no source at all.
+  // `sys-approval-display-title.test.ts` holds the two to the same text where
+  // no label is stored, and pins where they differ.
   displayNameField: 'display_title',
   nameField: 'display_title', // [ADR-0079] canonical primary-title pointer (mirrors deprecated displayNameField)
   titleFormat: '{process_name} · {record_id}',
@@ -128,12 +135,15 @@ export const SysApprovalRequest = ObjectSchema.create({
     id: Field.text({ label: 'Request ID', required: true, readonly: true, group: 'System' }),
 
     // [ADR-0079] The record title (`nameField` above). A formula is computed on
-    // read and has no stored column. Both source columns are required, so the
-    // expression needs no null guard.
+    // read and has no stored column. Its flow half is the flow's label,
+    // `process_label`, the same column the request lists show. That column is
+    // nullable — a row the boot fill could not label, or one written before
+    // the fill ran — so the expression guards it and falls back to
+    // `process_name`, which is required, as is `record_id`.
     display_title: Field.formula({
       label: 'Title',
       returnType: 'text',
-      expression: F`record.process_name + ' · ' + record.record_id`,
+      expression: F`record.process_label != null ? record.process_label + ' · ' + record.record_id : record.process_name + ' · ' + record.record_id`,
       description: 'Record title: the request source and the record it is about (computed on read)',
       group: 'Target',
     }),