fix(slugs): CJK slug support in SlugRegistry and dream-cycle summary slug (takeover of #782, #738)

Master already widened slugifySegment (sync.ts) and validatePageSlug
(operations.ts) to CJK in v0.32.7, but the other two validators #782
targeted stayed ASCII-only: SlugRegistry's SLUG_RE rejected any CJK
desiredSlug from BrainWriter, and synthesize.ts's SUMMARY_SLUG_RE (whose
comment claimed it was kept in sync with validatePageSlug) rejected CJK
output roots.

Hoist the segment grammar into cjk.ts as PAGE_SLUG_SEG and compose all
three regex sites from it, so the four slug validators share one grammar.
Each site keeps its own shape (SlugRegistry's >=2-segment dir/name form,
validatePageSlug's case-insensitive flag).

Scope stays CJK (matching v0.32.7), not full \p{L} Unicode as #782
proposed — all-scripts slugs (lookalike/RTL spoofing) is a maintainer
policy call.

Co-authored-by: tamagodo-fu <tamagodo-fu@users.noreply.github.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

src/commands/extract.ts

@@ -1743,15 +1743,7 @@ async function extractStaleFromDB(
       // `page.updated_at.toISOString()` — the JS Date is ms-truncated, so the
       // µs-precision DB updated_at stayed strictly greater and the page never
       // cleared on Postgres. Stamping the exact value makes them equal.
-      //
-      // Version-arm floor: a page last edited BEFORE LINK_EXTRACTOR_VERSION_TS
-      // would otherwise be stamped below the version watermark and stay
-      // permanently stale (`links_extracted_at < versionTs` re-fires every run).
-      // Stamp max(updated_at, versionTs) — versionTs is always a past release
-      // date, so a concurrent edit's now() still exceeds the stamp and D4 holds.
-      // Tie at ms precision picks updated_at_iso (its µs ≥ versionTs's .000000).
-      const stampTs = new Date(page.updated_at_iso) >= new Date(versionTs) ? page.updated_at_iso : versionTs;
-      processedRefs.push({ slug: page.slug, source_id: page.source_id, extractedAt: stampTs });
+      processedRefs.push({ slug: page.slug, source_id: page.source_id, extractedAt: page.updated_at_iso });
     }
 
     // Flush NON-swallowing (CDX-4): a throw here propagates out of the sweep so

src/core/cjk.ts

@@ -20,6 +20,14 @@ export const CJK_SLUG_CHARS = '一-鿿぀-ゟ゠-ヿ가-힯';
 
 export const CJK_RANGES_REGEX = new RegExp(`[${CJK_SLUG_CHARS}]`);
 
+/**
+ * Page-slug segment grammar (no anchors): alnum-or-CJK lead char, then
+ * alnum/CJK/hyphen continuation. Single source for validatePageSlug
+ * (operations.ts), SlugRegistry's SLUG_RE, and the dream-cycle
+ * SUMMARY_SLUG_RE so every slug validator shares one grammar (#738).
+ */
+export const PAGE_SLUG_SEG = `[a-z0-9${CJK_SLUG_CHARS}][a-z0-9${CJK_SLUG_CHARS}\\-]*`;
+
 export const CJK_SENTENCE_DELIMITERS = ['。', '！', '？']; // 。！？
 export const CJK_CLAUSE_DELIMITERS = ['；', '：', '，', '、']; // ；：，、
 

src/core/cycle/synthesize.ts

@@ -43,10 +43,11 @@ import { serializeMarkdown, serializePageToMarkdown } from '../markdown.ts';
 import type { Page, PageType } from '../types.ts';
 import { validateSourceId } from '../utils.ts';
 import { safeSplitIndex } from '../text-safe.ts';
+import { PAGE_SLUG_SEG } from '../cjk.ts';
 
-// Slug regex from validatePageSlug — kept in sync.
+// Slug grammar from validatePageSlug — shared via PAGE_SLUG_SEG (#738).
 // Used for the orchestrator-written summary index slug.
-const SUMMARY_SLUG_RE = /^[a-z0-9][a-z0-9\-]*(\/[a-z0-9][a-z0-9\-]*)*$/;
+const SUMMARY_SLUG_RE = new RegExp(`^${PAGE_SLUG_SEG}(\\/${PAGE_SLUG_SEG})*$`);
 
 // ── Model context budget (D1, D5, D7, D9) ─────────────────────────────
 

src/core/link-extraction.ts

@@ -28,7 +28,7 @@ import { ensureWellFormed } from './text-safe.ts';
  * OR updated_at > links_extracted_at`. It is an ISO-8601 string (NOT a number) —
  * the column is TIMESTAMPTZ and the predicate binds it as `::timestamptz`.
  */
-export const LINK_EXTRACTOR_VERSION_TS = '2026-07-21T00:00:00Z';
+export const LINK_EXTRACTOR_VERSION_TS = '2026-05-31T00:00:00Z';
 
 // ─── Entity references ──────────────────────────────────────────
 
@@ -80,10 +80,10 @@ export type LinkResolutionType = 'qualified' | 'unqualified';
  * Directory prefix whitelist. These are the top-level slug dirs the extractor
  * recognizes as entity references. Upstream canonical + our extensions:
  *   - Gbrain canonical: people, companies, meetings, concepts, deal, civic, project, source, media, yc, projects
- *   - Our domain extensions: tech, finance, personal, openclaw, ops (domain-organized wikis)
+ *   - Our domain extensions: tech, finance, personal, openclaw (domain-organized wikis)
  *   - Our entity prefix: entities (we kept some legacy entities/projects/ pages)
  */
-const DIR_PATTERN = '(?:people|companies|meetings|concepts|deal|civic|project|projects|source|media|yc|tech|finance|personal|openclaw|entities|ops)';
+const DIR_PATTERN = '(?:people|companies|meetings|concepts|deal|civic|project|projects|source|media|yc|tech|finance|personal|openclaw|entities)';
 
 /**
  * Match `[Name](path)` markdown links pointing to entity directories.
@@ -865,16 +865,7 @@ export function queryBasenameIndex(idx: Map<string, string[]>, name: string): st
   if (!name || typeof name !== 'string') return [];
   const trimmed = name.trim();
   if (!trimmed) return [];
-  let hit = idx.get(trimmed) ?? idx.get(trimmed.toLowerCase()) ?? idx.get(normalizeBasename(trimmed));
-  // Issue #2576 bug 2: path-style refs (`runbooks/2026-05-01-x`) from dirs
-  // outside DIR_PATTERN reach here, but normalizeBasename strips slashes
-  // into a garbage key (`runbooks2026-05-01-x`) that can never hit the
-  // tail-keyed index. Fall back to the path tail so qualified refs resolve
-  // by basename like everything else.
-  if (!hit && trimmed.includes('/')) {
-    const tail = trimmed.slice(trimmed.lastIndexOf('/') + 1).trim();
-    if (tail) hit = idx.get(tail) ?? idx.get(tail.toLowerCase()) ?? idx.get(normalizeBasename(tail));
-  }
+  const hit = idx.get(trimmed) ?? idx.get(trimmed.toLowerCase()) ?? idx.get(normalizeBasename(trimmed));
   return hit ? [...hit].sort(basenameSort) : [];
 }
 

src/core/operations.ts

@@ -25,7 +25,7 @@ import { bumpLastRetrievedAt } from './last-retrieved.ts';
 import { isSearchMode } from './search/mode.ts';
 import { stampEvidence } from './search/evidence.ts';
 import type { SearchResult } from './types.ts';
-import { CJK_SLUG_CHARS } from './cjk.ts';
+import { CJK_SLUG_CHARS, PAGE_SLUG_SEG } from './cjk.ts';
 import * as db from './db.ts';
 import { VERSION } from '../version.ts';
 import {
@@ -162,7 +162,6 @@ export function validatePageSlug(slug: string): void {
   }
   // v0.32.7: CJK ranges (Han / Hiragana / Katakana / Hangul Syllables) allowed
   // in segments. ASCII shape rules (lead char, hyphen continuation) preserved.
-  const PAGE_SLUG_SEG = `[a-z0-9${CJK_SLUG_CHARS}][a-z0-9${CJK_SLUG_CHARS}\\-]*`;
   if (!new RegExp(`^${PAGE_SLUG_SEG}(\\/${PAGE_SLUG_SEG})*$`, 'i').test(slug)) {
     throw new OperationError('invalid_params', `Invalid page_slug: ${slug} (allowed: alphanumeric, CJK, hyphens, forward-slash separated segments)`);
   }

src/core/output/slug-registry.ts

@@ -17,6 +17,7 @@
 
 import type { BrainEngine } from '../engine.ts';
 import type { PageType } from '../types.ts';
+import { PAGE_SLUG_SEG } from '../cjk.ts';
 
 export interface CreateSlugInput {
   /**
@@ -71,7 +72,9 @@ export class SlugRegistryError extends Error {
 // SlugRegistry
 // ---------------------------------------------------------------------------
 
-const SLUG_RE = /^[a-z0-9][a-z0-9\-]*(\/[a-z0-9][a-z0-9\-]*)+$/;
+// Shares the page-slug segment grammar (incl. CJK ranges, #738) with
+// validatePageSlug; keeps this site's dir/name shape (>= 2 segments).
+const SLUG_RE = new RegExp(`^${PAGE_SLUG_SEG}(\\/${PAGE_SLUG_SEG})+$`);
 
 export class SlugRegistry {
   constructor(private engine: BrainEngine) {}

test/cycle-dream-output-root.test.ts

@@ -86,6 +86,12 @@ describe('#2415: loadOutputRoot validation + patterns gather scope', () => {
     expect(await loadOutputRoot(engine)).toBe('wiki');
   });
 
+  test('CJK root passes the slug grammar (#738)', async () => {
+    await engine.setConfig('dream.synthesize.output_root', '知识/笔记');
+    expect(await loadOutputRoot(engine)).toBe('知识/笔记');
+    await engine.setConfig('dream.synthesize.output_root', '');
+  });
+
   test('patterns phase gathers reflections under the configured root', async () => {
     await engine.setConfig('dream.synthesize.output_root', 'notes');
     for (let i = 0; i < 3; i++) {

test/extract-stale.test.ts

@@ -209,24 +209,6 @@ describe('gbrain extract --stale', () => {
     expect(usRows[0]?.eq).toBe(true);
   });
 
-  test('version-arm floor: page edited BEFORE LINK_EXTRACTOR_VERSION_TS clears after --stale (issue #2576 bug 3)', async () => {
-    // A page whose updated_at predates the version watermark used to be
-    // stamped at its updated_at (< versionTs), so the version arm re-fired
-    // every run — permanently stale. The sweep now floors the stamp at
-    // versionTs. (The #1768 test above also covers this since the v0.42.x
-    // VERSION_TS bump moved its date below the watermark, but this pins the
-    // behavior explicitly so a date "repair" there can't drop coverage.)
-    await engine.putPage('people/alice', personPage('Alice'));
-    await engine.executeRaw(`UPDATE pages SET updated_at = '2000-01-01T00:00:00Z' WHERE slug = 'people/alice'`);
-    expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(1);
-
-    await runExtract(engine, ['--stale']);
-    // Pre-floor this stayed 1 forever (stamp < versionTs → version arm re-fires).
-    expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(0);
-    await runExtract(engine, ['--stale']);
-    expect(await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS })).toBe(0);
-  });
-
   test('CDX-4 (D2): a link-flush throw aborts the sweep and leaves pages UNSTAMPED', async () => {
     await engine.putPage('people/alice', personPage('Alice'));
     await engine.putPage('companies/acme', companyPage('Acme', '[Alice](people/alice) founded [Acme](companies/acme).'));

test/link-extraction.test.ts

@@ -140,15 +140,6 @@ describe('extractEntityRefs', () => {
     expect(wikiRefs[0].needsResolution).toBe(true);
   });
 
-  test('recognizes ops/ qualified wikilinks (issue #2576 bug 2)', () => {
-    // `ops` was missing from DIR_PATTERN, so [[ops/...]] fell through to
-    // the generic 2c pass (needsResolution) instead of being a real ref.
-    const refs = extractEntityRefs('Deployed via [[ops/services/pointer-agent]].');
-    expect(refs.length).toBe(1);
-    expect(refs[0].slug).toBe('ops/services/pointer-agent');
-    expect(refs[0].needsResolution).toBeUndefined();
-  });
-
   test('skips qualified-syntax tokens (those belong to 2a)', () => {
     // [[wiki:topics/ai]] looks like 2a's qualified shape — even though
     // it wouldn't satisfy DIR_PATTERN, 2c must not claim it either
@@ -1078,15 +1069,6 @@ describe('makeResolver — fallback chain', () => {
     ]);
   });
 
-  test('resolveBasenameMatches: path-style ref falls back to the tail (issue #2576 bug 2)', async () => {
-    // normalizeBasename strips slashes, so `runbooks/2026-05-01-pointer-agent`
-    // used to normalize to a garbage key that never hit the tail-keyed index.
-    const engine = makeFakeEngineWithSlugs(['ops/changes/2026-05-01-pointer-agent']);
-    const r = makeResolver(engine);
-    expect(await r.resolveBasenameMatches!('runbooks/2026-05-01-pointer-agent'))
-      .toEqual(['ops/changes/2026-05-01-pointer-agent']);
-  });
-
   test('resolveBasenameMatches: case-insensitive fallback', async () => {
     const engine = makeFakeEngineWithSlugs(['companies/fast-weigh']);
     const r = makeResolver(engine);

test/writer.test.ts

@@ -203,6 +203,17 @@ describe('SlugRegistry', () => {
     })).rejects.toThrow(SlugRegistryError);
   });
 
+  test('create accepts CJK slugs (#738)', async () => {
+    const reg = new SlugRegistry(engine);
+    const r = await reg.create({
+      desiredSlug: '知识/品牌圣经',
+      displayName: '品牌圣经',
+      type: 'note',
+    });
+    expect(r.slug).toBe('知识/品牌圣经');
+    expect(r.exact).toBe(true);
+  });
+
   test('create throws on invalid slug', async () => {
     const reg = new SlugRegistry(engine);
     await expect(reg.create({