@@ -0,0 +1,18 @@
+---
+'@objectstack/spec': patch
+'@objectstack/rest': patch
+'@object-ui/console': patch
+---
+
+Clause-②: no
+Metadata-protocol: none
+
+fix(spec): a book's *Uncategorized* group no longer lists the pages a sibling book places (#27091)
+
+When two books split one package's docs, for example a user manual and an admin manual gated by `audience: { permissionSet }`, each book's tree used to list the other book's pages under a trailing *Uncategorized* group. `resolveBookTree` judged "claimed" against the book's own groups only. ADR-0046 §6.4 rolls up only the docs "left unplaced", and a doc a sibling book places is placed.
+
+- **`resolveBookTree(book, docs, bookPackage?, books?)`** takes an optional fourth argument, every registered book. With it, the *Uncategorized* group holds only the docs of the book's packages that no book claims. The claim judge is `resolveBookClaimedDocs` over each book, the same claim set `resolveDocAudiences` unions audiences over. Without it, the tree is the same as before.
+- **`GET /meta/book/:name/tree`** passes the books it already reads, so an administrator opening the user manual no longer sees the admin pages under *Uncategorized*. A doc that no book claims still rolls up there.
+- **The console's docs pages** render the book sidebar with the console's own copy of the resolver (`book-nav.ts`). It takes the same optional fourth argument, and the reader and the book landing pass the authored books, so the sidebar no longer lists a sibling book's pages under *Uncategorized* either.
+
+The change only removes entries from the catch-all: a sibling book's docs appear nowhere new, and a doc's effective audience (§6.7) does not change.
@@ -396,7 +396,9 @@ export interface DocsAudience {
     /**
      * `book` resolved over `docs`, narrowed to the entries this caller may read
      * — exactly the body `GET /meta/book/:name/tree` serves once the book's own
-     * audience has admitted the caller. `canRead` is a {@link docReader} over
+     * audience has admitted the caller. Its *Uncategorized* group holds only
+     * the docs NO book of this audience claims (#27091), so a sibling book's
+     * pages never land there. `canRead` is a {@link docReader} over
      * the same `docs` when the caller already holds one.
      */
     readableTree(
@@ -702,7 +704,10 @@ export async function resolveDocsAudience(
         admitsBook: (book: Book) => audienceAllows(book?.audience, caller),
         docReader,
         readableTree: (book, docs, canRead) => {
-            const tree = resolveBookTree(book, docs, book._packageId);
+            // [#27091] Every book this audience was built over judges the
+            // *Uncategorized* catch-all, so a doc a sibling book places is not
+            // this book's orphan (ADR-0046 §6.4).
+            const tree = resolveBookTree(book, docs, book._packageId, books);
             // The fast path serves the tree whole — no entry can fail an
             // audience every doc passes, and the empty-group drop below is
             // part of the narrowing, not of the tree.
 
@@ -749,8 +749,9 @@ describe('RestServer', () => {
       const tree = res.json.mock.calls.at(-1)![0];
       const docsInTree = tree.groups.flatMap((g: any) => g.entries.map((e: any) => e.doc));
       expect(docsInTree).toContain('pub_intro');
-      // org_intro / gated_setup are claimed by other books; the orphan pass
-      // would render them, but the audience filter must drop them for anon.
+      // org_intro / gated_setup are claimed by other books, so they are not
+      // this book's orphans (#27091); the audience filter would drop them for
+      // anon even if they were.
       expect(docsInTree).not.toContain('org_intro');
       expect(docsInTree).not.toContain('gated_setup');
     });
@@ -784,6 +785,27 @@ describe('RestServer', () => {
       expect(docsInTree).toContain('gated_setup');
     });
 
+    it('tree: [#27091] a holder of every book sees no sibling book\'s pages under Uncategorized', async () => {
+      // Three books split one package's docs: each doc is claimed by exactly
+      // one book, so no tree has a doc "left unplaced" (ADR-0046 §6.4). The
+      // holder reads every doc, so only the resolver can keep them out.
+      protocol.getMetaItems = metaByType([...BOOKS], [...DOCS, { name: 'loose_note', _packageId: 'pkg' }]);
+      const rest = new RestServer(server as any, protocol as any, ANON_API as any);
+      (rest as any).resolveExecCtx = vi.fn().mockResolvedValue({ userId: 'u1' });
+      (rest as any).securityServiceProvider = async () => ({
+        resolvePermissionSetNames: async () => ['member_default', 'crm_admin'],
+      });
+      rest.registerRoutes();
+      const route = getMetaRoute(rest, 'GET', '/api/v1/meta/book/:name/tree');
+      for (const [name, own] of [['org_book', 'org_intro'], ['gated_book', 'gated_setup']]) {
+        const res = mockRes();
+        await route!.handler({ params: { name }, query: {}, headers: {} }, res);
+        const tree = res.json.mock.calls.at(-1)![0];
+        expect(Object.fromEntries(tree.groups.map((g: any) => [g.key, g.entries.map((e: any) => e.doc)])))
+          .toEqual({ g: [own], uncategorized: ['loose_note'] });
+      }
+    });
+
     it('list: anonymous /meta/book returns only public books', async () => {
       protocol.getMetaItems = metaByType();
       const rest = new RestServer(server as any, protocol as any, ANON_API as any);
 
@@ -47,7 +47,7 @@
     "tags": {
       "status": "live",
       "verifiedAt": "2026-09-27",
-      "evidence": "packages/spec/src/system/book.zod.ts#matchesInclude (`Array.isArray(doc.tags) && doc.tags.includes(include.tag)` — the tag branch of `include`); packages/spec/src/system/book.zod.ts#resolveBookTree (the caller that decides membership from it); packages/rest/src/meta-item-read-gate.ts#createMetaBookTreeAnswer (`tags: d.tags` in the tree corpus the route hands `audience.readableTree(book, docs)` — the transport half); packages/rest/src/meta-item-read-gate.ts#resolveBookTree (`resolveBookTree(book, docs, book._packageId)` — where that corpus reaches the resolver)",
+      "evidence": "packages/spec/src/system/book.zod.ts#matchesInclude (`Array.isArray(doc.tags) && doc.tags.includes(include.tag)` — the tag branch of `include`); packages/spec/src/system/book.zod.ts#resolveBookTree (the caller that decides membership from it); packages/rest/src/meta-item-read-gate.ts#createMetaBookTreeAnswer (`tags: d.tags` in the tree corpus the route hands `audience.readableTree(book, docs)` — the transport half); packages/rest/src/meta-item-read-gate.ts#resolveBookTree (`resolveBookTree(book, docs, book._packageId, books)` — where that corpus reaches the resolver)",
       "note": "DECLARED 2026-08-02 (#4509, ADR-0049) — the enforce half of enforce-or-remove. The consumer, the transport and the resolver-side interface all predated this key: matchesInclude compared `doc.tags` against a group's `include: { tag }`, the book route already forwarded `tags: d.tags`, and ResolverDoc declared `tags?: string[]` marked '(P3d; absent today)'. What was missing was one line HERE — DocSchema is strict, so authoring `tags:` was a parse error, every doc reached the resolver with tags undefined, and the `{ tag }` include variant could never match. Removing the variant was the alternative and was rejected: a union member has no clean tombstone (retiredKey covers object keys), so authors would have gotten a bare union error, and it would have discarded working matcher code. Live on arrival — the branch it feeds is reachable the moment a doc carries a tag. 2026-08-28: RE-ANCHORED (commit 8cb96ec41) and REPOINTED — `book.zod.ts:213` is `description?: string;` in the `ResolvedEntry` interface, and the parenthetical claiming it was \"matchesInclude tag branch\" named the right function ~61 lines away; the \"reached from :255 and :281\" hops were prose the ledger never bounded. `rest-server.ts:3218` had rotted onto `this.registerSecurityEndpoints(bp)` — a route-registration call, ~1,700 lines from the corpus that forwards the tags. Re-closed by hand against 93ea19bca."
     }
   }
 
@@ -262,6 +262,84 @@ describe('resolveBookTree — the Uncategorized group holds only the book\'s own
   });
 });
 
+// [#27091] ADR-0046 §6.4 rolls up docs "left unplaced"; a doc a SIBLING book
+// places is placed. Two books over one package (split by audience, §6.7) used
+// to list each other's pages under *Uncategorized*, because "claimed" was
+// judged against the book's own groups only.
+describe('resolveBookTree — a doc a sibling book claims is not this book\'s orphan (§6.4)', () => {
+  const corpus: ResolverDoc[] = [
+    { name: 'crm_overview', packageId: 'crm' },
+    { name: 'crm_sales_leads', packageId: 'crm' },
+    { name: 'crm_admin_access', packageId: 'crm' },
+    { name: 'crm_admin_automation', packageId: 'crm' },
+  ];
+  const withStray: ResolverDoc[] = [...corpus, { name: 'crm_stray', packageId: 'crm' }];
+  const userManual: Book & { packageId?: string } = {
+    name: 'crm_manual',
+    packageId: 'crm',
+    groups: [
+      { key: 'start', label: 'Start', include: 'crm_overview' },
+      { key: 'sales', label: 'Sales', include: 'crm_sales_*' },
+    ],
+  };
+  const adminManual: Book & { packageId?: string } = {
+    name: 'crm_admin_manual',
+    packageId: 'crm',
+    audience: { permissionSet: 'system_admin' },
+    groups: [{ key: 'admin', label: 'Admin', include: 'crm_admin_*' }],
+  };
+  const books = [userManual, adminManual];
+  const keysAndDocs = (tree: ResolvedBook) =>
+    Object.fromEntries(tree.groups.map((g) => [g.key, g.entries.map((e) => e.doc)]));
+  const docsOf = (tree: ResolvedBook) => tree.groups.flatMap((g) => g.entries.map((e) => e.doc));
+
+  it('two books with disjoint includes over one package: neither tree has an Uncategorized group', () => {
+    expect(keysAndDocs(resolveBookTree(userManual, corpus, 'crm', books))).toEqual({
+      start: ['crm_overview'],
+      sales: ['crm_sales_leads'],
+    });
+    expect(keysAndDocs(resolveBookTree(adminManual, corpus, 'crm', books))).toEqual({
+      admin: ['crm_admin_access', 'crm_admin_automation'],
+    });
+  });
+
+  it('a doc NO book claims still rolls up under Uncategorized, in every book of its package', () => {
+    expect(keysAndDocs(resolveBookTree(userManual, withStray, 'crm', books)).uncategorized).toEqual(['crm_stray']);
+    expect(keysAndDocs(resolveBookTree(adminManual, withStray, 'crm', books)).uncategorized).toEqual(['crm_stray']);
+  });
+
+  it('an audience-gated sibling\'s docs never reach the other book\'s tree, and nothing appears anywhere new', () => {
+    const userDocs = docsOf(resolveBookTree(userManual, withStray, 'crm', books));
+    expect(userDocs).not.toContain('crm_admin_access');
+    expect(userDocs).not.toContain('crm_admin_automation');
+    // The catch-all only shrinks: every entry with siblings was an entry without them.
+    for (const book of books) {
+      const before = docsOf(resolveBookTree(book, withStray, 'crm'));
+      expect(before).toEqual(expect.arrayContaining(docsOf(resolveBookTree(book, withStray, 'crm', books))));
+    }
+  });
+
+  it('the §6.7 audience union is unmoved: the claim sets never read the catch-all', () => {
+    expect(Object.fromEntries(resolveDocAudiences(books, withStray))).toEqual({
+      crm_overview: ['org'],
+      crm_sales_leads: ['org'],
+      crm_admin_access: [{ permissionSet: 'system_admin' }],
+      crm_admin_automation: [{ permissionSet: 'system_admin' }],
+      crm_stray: ['org'],
+    });
+    for (const book of books) {
+      const authored = resolveBookTree(book, withStray, 'crm', books).groups.filter((g) => g.key !== 'uncategorized');
+      expect([...resolveBookClaimedDocs(book, withStray, 'crm')]).toEqual(authored.flatMap((g) => g.entries.map((e) => e.doc)));
+    }
+  });
+
+  it('a single-book package is byte-identical to the tree without `books`', () => {
+    const alone = resolveBookTree(userManual, withStray, 'crm');
+    expect(JSON.stringify(resolveBookTree(userManual, withStray, 'crm', [userManual]))).toBe(JSON.stringify(alone));
+    expect(keysAndDocs(alone).uncategorized).toEqual(['crm_admin_access', 'crm_admin_automation', 'crm_stray']);
+  });
+});
+
 describe('deriveImplicitPackageBook + audience', () => {
   it('synthesizes a one-group book including every doc of the package', () => {
     const book = deriveImplicitPackageBook('app_todo', 'Todo');
 
@@ -302,18 +302,33 @@ function entryFromDoc(doc: ResolverDoc): ResolvedEntry {
  *  - Otherwise membership is derived: a doc joins the first group (in group
  *    order) whose `include` matches it OR whose `key` equals the doc's explicit
  *    `group`. Within a group, docs sort by `doc.order` then label.
- *  - Any doc OF THE BOOK'S PACKAGES claimed by no group falls into a
- *    synthetic *Uncategorized* group appended last — no doc of the book is
- *    ever dropped (ADR-0046 §6.4: "any doc left unplaced rolls up under a
- *    synthetic *Uncategorized* group", the unplaced docs being the
+ *  - Any doc OF THE BOOK'S PACKAGES that NO registered book claims falls
+ *    into a synthetic *Uncategorized* group appended last — no doc of the
+ *    book is ever dropped (ADR-0046 §6.4: "any doc left unplaced rolls up
+ *    under a synthetic *Uncategorized* group", the unplaced docs being the
  *    package's). The book's packages are `bookPackage` plus every group's
  *    `package`; a doc of any OTHER package is not this book's to catch — it
  *    stays reachable through its own package's book. When the book declares
  *    no package at all, every unclaimed doc is an orphan, as before.
  *    Explicit placement is NOT scoped: a doc of another package whose
  *    `group` names one of this book's groups still joins that group.
+ *  - "Claimed" is judged against `books`, every registered book (#27091):
+ *    a doc a SIBLING book places is placed, not "left unplaced" (§6.4), so
+ *    it is not this book's orphan — two books splitting one package's docs
+ *    (by audience, §6.7) no longer list each other's pages under
+ *    *Uncategorized*. The judge is `resolveBookClaimedDocs` over each book,
+ *    the very claim set `resolveDocAudiences` unions audiences over, so the
+ *    catch-all only ever SHRINKS: a sibling's docs leave it and appear
+ *    nowhere new. That claim set excludes the *Uncategorized* group, so the
+ *    §6.7 audience union cannot move. Without `books` only this book's own
+ *    groups judge, as before.
  */
-export function resolveBookTree(book: Book, docs: ResolverDoc[], bookPackage?: string): ResolvedBook {
+export function resolveBookTree(
+  book: Book,
+  docs: ResolverDoc[],
+  bookPackage?: string,
+  books?: readonly AudienceBook[],
+): ResolvedBook {
   const groupsSorted = [...book.groups]
     .map((g, i) => ({ g, i }))
     .sort((a, b) => (a.g.order ?? 0) - (b.g.order ?? 0) || a.i - b.i)
@@ -398,8 +413,9 @@ export function resolveBookTree(book: Book, docs: ResolverDoc[], bookPackage?: s
   const ownPackages = [bookPackage, ...book.groups.map((g) => g.package)].filter(
     (p): p is string => typeof p === 'string' && p.length > 0,
   );
+  const claimedByAnyBook = books ? claimedByBooks(books, docs) : undefined;
   const orphans = docs
-    .filter((d) => !claimed.has(d.name))
+    .filter((d) => !claimed.has(d.name) && !claimedByAnyBook?.has(d.name))
     .filter((d) => ownPackages.length === 0 || ownPackages.some((p) => matchesInclude(d, '*', p)))
     .sort(byOrderThenLabel);
   if (orphans.length) {
@@ -493,6 +509,26 @@ export interface AudienceBook extends Book {
   packageId?: string;
 }
 
+/**
+ * THE cross-book claim pass: each book with the docs it claims
+ * (`resolveBookClaimedDocs` over its own `packageId`). `resolveDocAudiences`
+ * unions audiences over it and `resolveBookTree`'s orphan pass reads its
+ * union ({@link claimedByBooks}), so the two never judge "claimed" apart.
+ */
+function* claimsPerBook(
+  books: readonly AudienceBook[],
+  docs: ResolverDoc[],
+): Generator<[AudienceBook, Set<string>]> {
+  for (const book of books) yield [book, resolveBookClaimedDocs(book, docs, book.packageId)];
+}
+
+/** Every doc some book in `books` claims: the union of {@link claimsPerBook}. */
+function claimedByBooks(books: readonly AudienceBook[], docs: ResolverDoc[]): Set<string> {
+  const all = new Set<string>();
+  for (const [, claimed] of claimsPerBook(books, docs)) for (const name of claimed) all.add(name);
+  return all;
+}
+
 /**
  * Effective audience of every doc = the union over the books that claim it
  * (ADR-0046 §6.7). A doc claimed by no authored book falls back to `'org'` —
@@ -507,9 +543,9 @@ export function resolveDocAudiences(
 ): Map<string, BookAudience[]> {
   const audiences = new Map<string, BookAudience[]>();
   for (const doc of docs) audiences.set(doc.name, []);
-  for (const book of books) {
+  for (const [book, claimed] of claimsPerBook(books, docs)) {
     const audience: BookAudience = book.audience ?? 'org';
-    for (const name of resolveBookClaimedDocs(book, docs, book.packageId)) {
+    for (const name of claimed) {
       audiences.get(name)?.push(audience);
     }
   }
 
@@ -45,12 +45,12 @@ export function DocRefusal({ name, message }: { name?: string; message?: string
 export default function BookPage() {
   const { slug, appName } = useParams<{ slug: string; appName?: string }>();
   const { t } = useObjectTranslation();
-  const { books, docs, state } = useBookData();
+  const { books, authoredBooks, docs, state } = useBookData();
 
   const found = useMemo(() => books.find((b) => bookSlug(b) === slug), [books, slug]);
   const opensTo = useMemo(
-    () => (found ? firstDoc(resolveBookTree(found, docs)) : undefined),
-    [found, docs],
+    () => (found ? firstDoc(resolveBookTree(found, docs, undefined, authoredBooks)) : undefined),
+    [found, docs, authoredBooks],
   );
 
   if (state === 'loading') {
 
@@ -158,12 +158,14 @@ export default function DocPage() {
   // reading in, so its spine renders as a persistent left nav. Resolved from
   // the portal book set; degrades to no sidebar if the slug is unknown (e.g.
   // the backend serves no books yet).
-  const { books, docs: allDocs } = useBookData();
+  const { books, authoredBooks, docs: allDocs } = useBookData();
   const base = appName ? `/apps/${appName}/docs` : '/docs';
   const resolvedBook = useMemo<ResolvedBook | null>(() => {
     const book = books.find((b) => bookSlug(b) === slug);
-    return book ? resolveBookTree(book, allDocs) : null;
-  }, [books, allDocs, slug]);
+    // Every authored book judges the Uncategorized group, so a doc a sibling
+    // book places is not listed there (objectstack#27091).
+    return book ? resolveBookTree(book, allDocs, undefined, authoredBooks) : null;
+  }, [books, authoredBooks, allDocs, slug]);
   const docHref = useCallback((docName: string) => `${base}/${slug}/${docName}`, [base, slug]);
   // A `:slug` segment that is a book's NAME rather than its slug — what a
   // `{ type: 'doc', book, doc }` navigation entry links to (objectui#11197) — is
Original file line number	Diff line number	Diff line change
`@@ -47,7 +47,7 @@`
`47`	`47`	`"tags": {`
`48`	`48`	`"status": "live",`
`49`	`49`	`"verifiedAt": "2026-09-27",`
`50`		- "evidence": "packages/spec/src/system/book.zod.ts#matchesInclude (`Array.isArray(doc.tags) && doc.tags.includes(include.tag)` — the tag branch of `include`); packages/spec/src/system/book.zod.ts#resolveBookTree (the caller that decides membership from it); packages/rest/src/meta-item-read-gate.ts#createMetaBookTreeAnswer (`tags: d.tags` in the tree corpus the route hands `audience.readableTree(book, docs)` — the transport half); packages/rest/src/meta-item-read-gate.ts#resolveBookTree (`resolveBookTree(book, docs, book._packageId)` — where that corpus reaches the resolver)",
	`50`	+ "evidence": "packages/spec/src/system/book.zod.ts#matchesInclude (`Array.isArray(doc.tags) && doc.tags.includes(include.tag)` — the tag branch of `include`); packages/spec/src/system/book.zod.ts#resolveBookTree (the caller that decides membership from it); packages/rest/src/meta-item-read-gate.ts#createMetaBookTreeAnswer (`tags: d.tags` in the tree corpus the route hands `audience.readableTree(book, docs)` — the transport half); packages/rest/src/meta-item-read-gate.ts#resolveBookTree (`resolveBookTree(book, docs, book._packageId, books)` — where that corpus reaches the resolver)",
`51`	`51`	"note": "DECLARED 2026-08-02 (#4509, ADR-0049) — the enforce half of enforce-or-remove. The consumer, the transport and the resolver-side interface all predated this key: matchesInclude compared `doc.tags` against a group's `include: { tag }`, the book route already forwarded `tags: d.tags`, and ResolverDoc declared `tags?: string[]` marked '(P3d; absent today)'. What was missing was one line HERE — DocSchema is strict, so authoring `tags:` was a parse error, every doc reached the resolver with tags undefined, and the `{ tag }` include variant could never match. Removing the variant was the alternative and was rejected: a union member has no clean tombstone (retiredKey covers object keys), so authors would have gotten a bare union error, and it would have discarded working matcher code. Live on arrival — the branch it feeds is reachable the moment a doc carries a tag. 2026-08-28: RE-ANCHORED (commit 8cb96ec41) and REPOINTED — `book.zod.ts:213` is `description?: string;` in the `ResolvedEntry` interface, and the parenthetical claiming it was \"matchesInclude tag branch\" named the right function ~61 lines away; the \"reached from :255 and :281\" hops were prose the ledger never bounded. `rest-server.ts:3218` had rotted onto `this.registerSecurityEndpoints(bp)` — a route-registration call, ~1,700 lines from the corpus that forwards the tags. Re-closed by hand against 93ea19bca."
`52`	`52`	`}`
`53`	`53`	`}`