fix(webhook): extract links for incremental push syncs (#2849)

GitHub webhook + sync-trigger 'sync' jobs never opted into extraction:
the standalone sync handler defaults noExtract to TRUE, so webhook-
imported pages advanced the source bookmark with links_extracted_at
unstamped — and nothing behind a standalone job ever swept them (the
autopilot cycle's extract phase only walks that cycle's changedSlugs).

Fix, both halves of the issue:
- webhook + sync-trigger payloads pass noExtract:false so normal
  incremental pushes get sync's inline link/timeline extraction.
- >100-file deferral branch: performSync surfaces extractDeferred, and
  the sync job handler queues a source-scoped extract{stale:true}
  follow-up job (idempotent sweep; extract handler gains a stale mode
  backed by the same extractStaleFromDB core as `gbrain extract --stale`).

Tests: new sync-large-deferral-extract.serial.test.ts drives a real
101-file incremental sync end-to-end (deferral flag → queued follow-up →
stale sweep recovers links + watermark); webhook payload contract test +
sync-trigger noExtract assertion salvaged from #2850.

Fixes #2849
Takeover of #2850

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

src/commands/autopilot.ts

@@ -38,7 +38,6 @@ import { logSelfUpgrade } from '../core/audit/self-upgrade-audit.ts';
 import { detectInstallMethod } from './upgrade.ts';
 import { evaluateQuietHours } from '../core/minions/quiet-hours.ts';
 import { inspectLock } from '../core/db-lock.ts';
-import { registerCleanup } from '../core/process-cleanup.ts';
 
 /**
  * v0.37.7.0 #1162 — classify autopilot reconnect-loop errors.
@@ -434,37 +433,6 @@ export async function runAutopilot(engine: BrainEngine, args: string[]) {
   let stopping = false;
   let childSupervisor: ChildWorkerSupervisor | null = null;
 
-  // #1872: graceful engine shutdown. On PGLite the cycle steps run INLINE in
-  // this process, so a hard `process.exit` mid-write (systemctl stop →
-  // SIGTERM) kills WASM Postgres with the WAL dirty and can corrupt the
-  // brain. Two exit paths must both close the engine:
-  //   - autopilot's own shutdown() below (owns SIGINT + internal stops like
-  //     max_crashes / cycle-failure-cap), and
-  //   - process-cleanup's SIGTERM handler (installed at cli.ts module load;
-  //     it runs the cleanup registry with a 3s deadline and then exits) —
-  //     which is why closeEngine is ALSO registered there.
-  // closeEngine aborts the in-flight inline cycle (runCycle checks the
-  // signal between phases and threads it into phase sub-work), gives it a
-  // short bounded window to wind down, then disconnects. PGLite's
-  // disconnect() drains the pending query and checkpoints before closing;
-  // a second call is a no-op (disconnect snapshots + nulls the handle), so
-  // both paths firing is safe.
-  const shutdownAbort = new AbortController();
-  let inflightInlineCycle: Promise<unknown> | null = null;
-  const closeEngine = async () => {
-    shutdownAbort.abort(new Error('autopilot shutdown'));
-    if (inflightInlineCycle) {
-      // ponytail: 2s cap keeps us inside process-cleanup's 3s deadline; a
-      // between-phase abort resolves instantly, a mid-phase one may not.
-      await Promise.race([
-        inflightInlineCycle.catch(() => { /* cycle errors already logged by the loop */ }),
-        new Promise((r) => setTimeout(r, 2_000)),
-      ]);
-    }
-    try { await engine.disconnect(); } catch { /* best-effort */ }
-  };
-  const deregisterEngineClose = registerCleanup('autopilot-engine-close', closeEngine);
-
   if (spawnManagedWorker) {
     const cliPath = resolveGbrainCliPath();
     // Cgroup-aware auto-sized RSS watchdog cap (issue #1678). The old flat
@@ -552,10 +520,6 @@ export async function runAutopilot(engine: BrainEngine, args: string[]) {
         childSupervisor.killChild('SIGKILL');
       }
     }
-    // #1872: abort the in-flight inline cycle and close the engine BEFORE
-    // process.exit — a hard exit mid-write corrupts PGLite's WASM Postgres.
-    await closeEngine();
-    deregisterEngineClose();
     try { unlinkSync(lockPath); } catch { /* already gone */ }
     process.exit(0);
   };
@@ -1044,21 +1008,16 @@ export async function runAutopilot(engine: BrainEngine, args: string[]) {
       // path's phase set). Now both converge on the same primitive.
       try {
         const { runCycle } = await import('../core/cycle.ts');
-        // #1872: track the promise so closeEngine can drain it on shutdown,
-        // and pass the abort signal so the cycle winds down between phases.
-        const cyclePromise = runCycle(engine, {
+        const report = await runCycle(engine, {
           brainDir: repoPath,
           // Autopilot daemon path: pulls by default (matches
           // pre-v0.17 autopilot behavior). CLI dream defaults false
           // for cron safety; that choice is scoped to dream only.
           pull: true,
-          signal: shutdownAbort.signal,
           yieldBetweenPhases: async () => {
             await new Promise(r => setImmediate(r));
           },
         });
-        inflightInlineCycle = cyclePromise;
-        const report = await cyclePromise.finally(() => { inflightInlineCycle = null; });
         // Only 'failed' (every attempted phase failed) trips the autopilot
         // circuit breaker. 'partial' means at least one phase warned or
         // failed while others ran — that's a soft signal, not a fatal

src/commands/dream.ts

@@ -26,7 +26,6 @@
 import type { BrainEngine } from '../core/engine.ts';
 import {
   runCycle,
-  resolveSourceForDir,
   ALL_PHASES,
   type CyclePhase,
   type CycleReport,
@@ -381,9 +380,9 @@ Options:
 
   --source <id>       Scope the cycle to one source so doctor's
                       cycle_freshness check sees a fresh stamp on
-                      completion. When omitted, gbrain derives the
-                      source from --dir / the configured checkout
-                      when it matches a source's local_path (#1869).
+                      completion. Without this, gbrain dream's
+                      timestamp never lands and federated brains
+                      see "stale cycle" forever.
   --source-id <id>    Alias for --source. Matches the v0.37.7.0+
                       naming used by import/extract/graph-query.
 
@@ -635,25 +634,6 @@ export async function runDream(engine: BrainEngine | null, args: string[]): Prom
     );
     process.exit(1);
   }
-
-  // #1869: a path-scoped run (--dir, or the configured sync.repo_path) whose
-  // directory matches a registered source's local_path IS that source's cycle
-  // — derive the source id so runCycle writes last_source_cycle_at /
-  // last_full_cycle_at on success and doctor's cycle_freshness check stops
-  // reading perpetually stale. Explicit --source still wins (resolved above).
-  // Fixed here at the command level, NOT in runCycle's stamp gate, so legacy
-  // global callers (autopilot-global-maintenance runs GLOBAL_PHASES with a
-  // brainDir and no sourceId) can't falsely stamp per-source freshness.
-  // A derived match on an archived source is skipped silently (falls back to
-  // legacy unscoped behavior) — stamping it would mask staleness on restore,
-  // mirroring the explicit --source archived guard above.
-  if (resolvedSourceId === undefined && engine !== null && brainDir !== null) {
-    const derived = await resolveSourceForDir(engine, brainDir);
-    if (derived !== undefined) {
-      const src = await fetchSource(engine, derived);
-      if (src?.archived !== true) resolvedSourceId = derived;
-    }
-  }
   // ─── issue #1678: bounded single-hold extract_atoms drain ──────────
   if (opts.drain) {
     if (engine === null) {

src/commands/extract.ts

@@ -1651,7 +1651,7 @@ async function extractTimelineFromDB(
  * make re-extraction idempotent). EVERY processed page is stamped, including
  * zero-link pages — they WERE processed.
  */
-async function extractStaleFromDB(
+export async function extractStaleFromDB(
   engine: BrainEngine,
   opts: {
     dryRun: boolean;

src/commands/jobs.ts

@@ -1479,7 +1479,31 @@ export async function registerBuiltinHandlers(
       embedSkipReason = 'auto_embed_disabled';
     }
 
-    return { ...result, embed_job_id: embedJobId, embed_skip_reason: embedSkipReason };
+    // #2849: large-sync extract deferral follow-up. performSync skips inline
+    // link/timeline extraction when totalChanges > 100, leaving
+    // links_extracted_at unstamped. A standalone sync job (webhook push,
+    // sync trigger) has no autopilot extract phase behind it, so the pages
+    // would stay extraction-stale until a manual `gbrain extract --stale`.
+    // Queue a source-scoped stale sweep instead. Best-effort + idempotent:
+    // a duplicate sweep finds 0 stale pages and no-ops.
+    let extractJobId: number | null = null;
+    if (result.extractDeferred) {
+      try {
+        const { MinionQueue } = await import('../core/minions/queue.ts');
+        const queue = new MinionQueue(engine);
+        const followUp = await queue.add(
+          'extract',
+          { stale: true, ...(sourceId ? { sourceId } : {}) },
+          {
+            idempotency_key: `sync-extract-stale:${sourceId ?? 'default'}:${Math.floor(Date.now() / 30_000)}`,
+            maxWaiting: 1,
+          },
+        );
+        extractJobId = followUp.id;
+      } catch { /* best-effort: extract --stale sweeps it later */ }
+    }
+
+    return { ...result, embed_job_id: embedJobId, embed_skip_reason: embedSkipReason, extract_stale_job_id: extractJobId };
   });
 
   registerBuiltinJob(worker, engine, 'embed', async (job) => {
@@ -1652,6 +1676,20 @@ export async function registerBuiltinHandlers(
   });
 
   worker.register('extract', async (job) => {
+    // #2849: stale-sweep mode — the sync handler's large-sync deferral
+    // follow-up. DB-source (reads page content from the DB, so it runs on
+    // checkout-less brains), source-scopable, idempotent. Same core as
+    // `gbrain extract --stale`.
+    if (job.data.stale === true) {
+      const { extractStaleFromDB } = await import('./extract.ts');
+      return await extractStaleFromDB(engine, {
+        dryRun: !!job.data.dryRun,
+        jsonMode: false,
+        includeFrontmatter: false,
+        sourceIdFilter: typeof job.data.sourceId === 'string' ? job.data.sourceId : undefined,
+        catchUp: false,
+      });
+    }
     const { runExtractCore } = await import('./extract.ts');
     const mode = (typeof job.data.mode === 'string' && ['links', 'timeline', 'all'].includes(job.data.mode))
       ? (job.data.mode as 'links' | 'timeline' | 'all')

src/commands/serve-http.ts

@@ -2146,8 +2146,13 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
   //     Other event types (ping, pull_request, etc.) return 202 'ignored'
   //     so GitHub doesn't retry.
   // D15.5: HMAC compare uses the shared safeHexEqual helper.
-  // D18: submits 'sync' job with auto_embed_backfill=true and priority -10
-  //     (above autopilot's 0).
+  // D18: submits 'sync' job with extraction + auto_embed_backfill enabled and
+  //     priority -10 (above autopilot's 0). noExtract:false opts normal
+  //     incremental pushes into sync's inline link/timeline extraction (#2849
+  //     — the standalone sync handler defaults noExtract to TRUE, which left
+  //     webhook-imported pages permanently stale). Large (>100 file) pushes
+  //     defer inline extract; the sync handler queues an extract --stale
+  //     follow-up job for that branch.
   // ---------------------------------------------------------------------------
   const githubWebhookLimiter = rateLimit({
     windowMs: 60_000,
@@ -2267,6 +2272,7 @@ export async function runServeHttp(engine: BrainEngine, options: ServeHttpOption
           'sync',
           {
             sourceId: source.id,
+            noExtract: false,
             auto_embed_backfill: true,
             embed_reason: 'webhook',
           },

src/commands/sync.ts

@@ -222,6 +222,14 @@ export interface SyncResult {
    * everything," the exact misdiagnosis in the #1794 recurrence report.
    */
   bankedFiles?: number;
+  /**
+   * #2849: true when extraction was REQUESTED (noExtract false) but this sync
+   * skipped inline link/timeline extraction because totalChanges > 100 (the
+   * #1794 large-sync deferral). links_extracted_at stays unstamped for the
+   * imported pages. The standalone `sync` job handler queues a source-scoped
+   * `extract --stale` follow-up when set; CLI runs print the manual hint.
+   */
+  extractDeferred?: boolean;
 }
 
 /**
@@ -1379,6 +1387,10 @@ See also:
     {
       sourceId: sourceIdArg,
       repoPath: source.local_path,
+      // #2849: opt in to inline extraction — the standalone sync handler
+      // defaults noExtract to TRUE (dedupe for doctor's [sync, extract]
+      // remediation plan), which would leave triggered syncs extraction-stale.
+      noExtract: false,
       auto_embed_backfill: true,
       embed_reason: 'sync_trigger',
     },
@@ -3287,11 +3299,16 @@ async function performSyncInner(engine: BrainEngine, opts: SyncOpts): Promise<Sy
   // the stale sweep scans the whole source, so banked-across-runs pages are
   // covered regardless.
   const extractOpts = opts.sourceId ? { sourceId: opts.sourceId } : undefined;
+  let extractDeferred = false;
   if (!opts.noExtract && totalChanges > 100 && pagesAffected.length > 0) {
+    // #2849: surface the deferral to callers. A standalone sync job (webhook
+    // push, sync trigger) has no autopilot extract phase behind it, so the
+    // job handler queues an `extract --stale` follow-up off this flag.
+    extractDeferred = true;
     slog(
       `  Large sync: deferring link/timeline extraction. ` +
       `Run 'gbrain extract --stale${opts.sourceId ? ` --source-id ${opts.sourceId}` : ''}' ` +
-      `(or let the autopilot cycle's extract phase sweep it).`,
+      `(sync jobs queue this follow-up automatically).`,
     );
   }
   if (!opts.noExtract && totalChanges <= 100 && pagesAffected.length > 0) {
@@ -3400,6 +3417,7 @@ async function performSyncInner(engine: BrainEngine, opts: SyncOpts): Promise<Sy
     chunksCreated,
     embedded,
     pagesAffected,
+    extractDeferred,
   };
 }
 

src/core/cycle.ts

@@ -855,16 +855,8 @@ interface SyncPhaseResult extends PhaseResult {
  * Resolve the source id for a brain directory by looking up the sources
  * table. Returns undefined when no registered source matches (falls back
  * to pre-v0.18 global config.sync.* keys).
- *
- * Exported for dream.ts (#1869): a `gbrain dream --dir <path>` run whose
- * path matches a registered source's local_path is a per-source cycle in
- * everything but name, so dream derives the source id up front and passes
- * it as opts.sourceId — landing the freshness stamp without changing
- * runCycle's stamp/lock semantics for legacy global callers (the
- * autopilot-global-maintenance handler runs GLOBAL_PHASES with a brainDir
- * and MUST NOT stamp per-source freshness; see rejected PR #2549).
  */
-export async function resolveSourceForDir(
+async function resolveSourceForDir(
   engine: BrainEngine,
   brainDir: string | null,
 ): Promise<string | undefined> {

test/autopilot-shutdown-engine-close.test.ts

@@ -1,63 +0,0 @@
-/**
- * #1872 — autopilot SIGTERM/SIGINT must close the engine before exit.
- *
- * On PGLite the cycle steps run INLINE in the autopilot process, so a hard
- * `process.exit` mid-write (systemctl stop → SIGTERM) kills WASM Postgres
- * with the WAL dirty and can corrupt the brain. Two exit paths must both
- * close the engine:
- *
- *   - autopilot's own shutdown() (owns SIGINT + internal stops like
- *     max_crashes / cycle-failure-cap), and
- *   - process-cleanup's SIGTERM handler (installed at cli.ts module load,
- *     which exits within its 3s cleanup deadline) — reached via the
- *     registered 'autopilot-engine-close' cleanup callback.
- *
- * Because the shutdown path is deep inside `runAutopilot()` (a long-running
- * daemon loop that ends in process.exit), a behavioral test would have to
- * spawn + signal a real daemon. Following the established precedent
- * (test/autopilot-supervisor-wiring.test.ts, test/autopilot-fanout-wiring.test.ts),
- * these static-shape regressions pin the load-bearing wiring instead.
- */
-import { describe, expect, it } from 'bun:test';
-import { readFileSync } from 'fs';
-import { join } from 'path';
-
-const AUTOPILOT_SRC = readFileSync(
-  join(import.meta.dir, '..', 'src', 'commands', 'autopilot.ts'),
-  'utf8',
-);
-
-describe('autopilot.ts graceful engine shutdown (#1872)', () => {
-  it('registers an engine-close callback in the process-cleanup registry (SIGTERM path)', () => {
-    // process-cleanup owns SIGTERM (installed at cli.ts:10) and hard-exits
-    // after its cleanup pass; without this registration the engine is never
-    // closed on `systemctl stop`.
-    expect(AUTOPILOT_SRC).toContain(
-      "import { registerCleanup } from '../core/process-cleanup.ts';",
-    );
-    expect(AUTOPILOT_SRC).toContain(
-      "registerCleanup('autopilot-engine-close', closeEngine)",
-    );
-  });
-
-  it('closeEngine aborts the in-flight inline cycle then disconnects the engine', () => {
-    // Abort first (runCycle checks the signal between phases and threads it
-    // into phase sub-work), bounded drain, then disconnect.
-    expect(AUTOPILOT_SRC).toMatch(
-      /const closeEngine = async \(\) => \{[\s\S]{0,900}shutdownAbort\.abort\([\s\S]{0,900}engine\.disconnect\(\)/,
-    );
-  });
-
-  it('the inline runCycle call carries the shutdown abort signal and is tracked as in-flight', () => {
-    // PGLite / --inline path: the cycle runs in-process, so shutdown must be
-    // able to (a) signal it to wind down and (b) await it before closing.
-    expect(AUTOPILOT_SRC).toMatch(/signal:\s*shutdownAbort\.signal/);
-    expect(AUTOPILOT_SRC).toMatch(/inflightInlineCycle\s*=\s*cyclePromise/);
-  });
-
-  it('shutdown() awaits closeEngine() before process.exit(0) (SIGINT + internal-stop path)', () => {
-    expect(AUTOPILOT_SRC).toMatch(
-      /await closeEngine\(\);[\s\S]{0,400}process\.exit\(0\)/,
-    );
-  });
-});

test/dream-dir-source-stamp.test.ts

@@ -1,99 +0,0 @@
-/**
- * #1869 — `gbrain dream --dir <path>` stamps cycle freshness when the path
- * matches a registered source's local_path.
- *
- * Pre-fix, only `--source <id>` runs wrote last_source_cycle_at /
- * last_full_cycle_at (runCycle's stamp gate reads opts.sourceId, and dream
- * never derived one from --dir), so a path-scoped brain showed doctor's
- * cycle_freshness as perpetually stale.
- *
- * The fix lives in dream.ts (derive the source id from the resolved brain
- * dir via resolveSourceForDir), NOT in runCycle's stamp gate — a runCycle-
- * wide change would make the autopilot-global-maintenance handler (global
- * phases, brainDir set, no sourceId) falsely stamp per-source freshness
- * (the #2194 poisoning class; see rejected PR #2549).
- *
- * Same real-PGLite/no-mocks discipline as test/dream.test.ts; same
- * GBRAIN_HOME isolation as test/cycle-last-full-cycle-at.test.ts (the
- * cycle's PGLite file lock lives under ~/.gbrain).
- */
-import { describe, test, expect, beforeAll, afterAll, beforeEach, afterEach } from 'bun:test';
-import { mkdtempSync, rmSync } from 'fs';
-import { join } from 'path';
-import { tmpdir } from 'os';
-import { PGLiteEngine } from '../src/core/pglite-engine.ts';
-import { resetPgliteState } from './helpers/reset-pglite.ts';
-import { runDream } from '../src/commands/dream.ts';
-import { withEnv } from './helpers/with-env.ts';
-
-let engine: PGLiteEngine;
-let brainDir: string;
-let gbrainHome: string;
-
-beforeAll(async () => {
-  engine = new PGLiteEngine();
-  await engine.connect({});
-  await engine.initSchema();
-}, 60_000);
-
-afterAll(async () => {
-  await engine.disconnect();
-});
-
-beforeEach(async () => {
-  await resetPgliteState(engine);
-  brainDir = mkdtempSync(join(tmpdir(), 'gbrain-dream-stamp-'));
-  gbrainHome = mkdtempSync(join(tmpdir(), 'gbrain-dream-stamp-home-'));
-}, 60_000);
-
-afterEach(() => {
-  rmSync(brainDir, { recursive: true, force: true });
-  rmSync(gbrainHome, { recursive: true, force: true });
-});
-
-async function seedSource(id: string, archived = false): Promise<void> {
-  await engine.executeRaw(
-    `INSERT INTO sources (id, name, local_path, config, archived, created_at)
-     VALUES ($1, $2, $3, '{}'::jsonb, $4, NOW())`,
-    [id, id, brainDir, archived],
-  );
-}
-
-async function readLastFullCycleAt(sourceId: string): Promise<string | null> {
-  const rows = await engine.executeRaw<{ config: Record<string, unknown> | null }>(
-    `SELECT config FROM sources WHERE id = $1`,
-    [sourceId],
-  );
-  const raw = rows[0]?.config?.last_full_cycle_at;
-  return typeof raw === 'string' ? raw : null;
-}
-
-describe('gbrain dream --dir <path> freshness stamp (#1869)', () => {
-  test('--dir matching a source local_path stamps last_full_cycle_at', async () => {
-    await withEnv({ GBRAIN_HOME: gbrainHome }, async () => {
-      await seedSource('path-scoped');
-      expect(await readLastFullCycleAt('path-scoped')).toBeNull();
-
-      const report = await runDream(engine, ['--dir', brainDir, '--phase', 'lint', '--json']);
-      expect(report).toBeTruthy();
-      if (report) expect(['ok', 'clean']).toContain(report.status);
-
-      // Pre-fix this stays null forever: dream never passed a sourceId, so
-      // runCycle's stamp gate skipped the write.
-      expect(await readLastFullCycleAt('path-scoped')).not.toBeNull();
-    });
-  }, 60_000);
-
-  test('--dir matching an ARCHIVED source does not stamp it', async () => {
-    await withEnv({ GBRAIN_HOME: gbrainHome }, async () => {
-      await seedSource('mothballed', true);
-
-      const report = await runDream(engine, ['--dir', brainDir, '--phase', 'lint', '--json']);
-      expect(report).toBeTruthy();
-
-      // Stamping an archived source would mask data staleness when it is
-      // later restored (mirrors the explicit --source archived guard).
-      expect(await readLastFullCycleAt('mothballed')).toBeNull();
-    });
-  }, 60_000);
-});

test/dream.test.ts

@@ -562,22 +562,12 @@ describe('runDream — --source / --source-id (v0.41.13)', () => {
 
   // ─── Back-compat: bare `gbrain dream` does NOT write per-source stamp ─
 
-  test('gbrain dream (no --source) stamps only the source whose local_path matches --dir (#1869)', async () => {
-    // Pre-#1869 this asserted NO source was ever stamped without an explicit
-    // --source — which is exactly the bug: a path-scoped `gbrain dream --dir`
-    // run never landed a freshness stamp and doctor's cycle_freshness stayed
-    // stale forever. New truth: the source whose local_path matches the
-    // resolved brain dir is derived and stamped; unrelated sources stay
-    // untouched (cross-source isolation).
-    await seedSource('alpha'); // local_path = repo → derived + stamped
-    await engine.executeRaw(
-      `INSERT INTO sources (id, name, local_path, config, archived, created_at)
-       VALUES ($1, $2, $3, '{}'::jsonb, false, NOW())`,
-      ['beta', 'beta', '/somewhere/else'],
-    );
+  test('gbrain dream (no --source) leaves all sources untouched (back-compat regression)', async () => {
+    await seedSource('alpha');
+    await seedSource('beta');
     const report = await runDream(engine, ['--dir', repo, '--phase', 'lint', '--json']);
     expect(report).toBeTruthy();
-    expect(await readLastFullCycleAt('alpha')).not.toBeNull();
+    expect(await readLastFullCycleAt('alpha')).toBeNull();
     expect(await readLastFullCycleAt('beta')).toBeNull();
   }, 60_000);
 

test/sources-webhook.test.ts

@@ -16,6 +16,7 @@
  */
 import { describe, test, expect } from 'bun:test';
 import { createHmac } from 'node:crypto';
+import { readFileSync } from 'node:fs';
 import { safeHexEqual } from '../src/core/timing-safe.ts';
 
 const GITHUB_SECRET = 'super-secret-webhook-key';
@@ -123,3 +124,25 @@ describe('Branch ref construction (D5)', () => {
     expect(pushedRef === `refs/heads/${trackedBranch}`).toBe(false);
   });
 });
+
+describe('Webhook sync job extraction contract (#2849)', () => {
+  test('opts into extraction before the pushed commit is consumed', () => {
+    const serveSource = readFileSync(
+      new URL('../src/commands/serve-http.ts', import.meta.url),
+      'utf8',
+    );
+    const routeStart = serveSource.indexOf("'/webhooks/github'");
+    const queueStart = serveSource.indexOf('const job = await queue.add(', routeStart);
+    const responseStart = serveSource.indexOf('res.status(202)', queueStart);
+    expect(routeStart).toBeGreaterThanOrEqual(0);
+    expect(queueStart).toBeGreaterThan(routeStart);
+    expect(responseStart).toBeGreaterThan(queueStart);
+
+    const routeSource = serveSource.slice(queueStart, responseStart);
+    const payload = routeSource.match(
+      /queue\.add\(\s*'sync',\s*\{([\s\S]*?)\}\s*,\s*\{/,
+    );
+    expect(payload).not.toBeNull();
+    expect(payload?.[1]).toMatch(/\bnoExtract:\s*false\b/);
+  });
+});

test/sync-large-deferral-extract.serial.test.ts

@@ -0,0 +1,130 @@
+/**
+ * #2849 — large-sync extract deferral queues an `extract --stale` follow-up.
+ *
+ * performSync's incremental path skips inline link/timeline extraction when
+ * totalChanges > 100 (the #1794 large-sync deferral), leaving
+ * links_extracted_at unstamped. Pre-fix, a standalone sync job (webhook push,
+ * `gbrain sync trigger`) had NOTHING behind it to sweep those pages — the
+ * autopilot cycle's extract phase only walks that cycle's changedSlugs — so a
+ * large webhook push left extraction permanently stale until a manual
+ * `gbrain extract --stale`.
+ *
+ * Pins:
+ *   (a) performSync surfaces `extractDeferred: true` on the >100 branch and
+ *       leaves the pages unstamped/unlinked.
+ *   (b) the `sync` job handler queues an `extract` job with
+ *       { stale: true, sourceId? } when extractDeferred is set.
+ *   (c) the `extract` handler's stale mode actually sweeps: links created +
+ *       watermark stamped (end-to-end recovery, no manual step).
+ *
+ * Marked .serial.test.ts — spawns git subprocesses + shares one PGLite engine.
+ */
+
+import { describe, test, expect, beforeAll, afterAll } from 'bun:test';
+import { mkdtempSync, writeFileSync, rmSync, mkdirSync } from 'fs';
+import { execSync } from 'child_process';
+import { tmpdir } from 'os';
+import { join } from 'path';
+import { PGLiteEngine } from '../src/core/pglite-engine.ts';
+import { MinionWorker } from '../src/core/minions/worker.ts';
+import { MinionQueue } from '../src/core/minions/queue.ts';
+import { registerBuiltinHandlers } from '../src/commands/jobs.ts';
+
+let engine: PGLiteEngine;
+let worker: MinionWorker;
+let repoPath: string;
+
+function git(cmd: string): void { execSync(cmd, { cwd: repoPath, stdio: 'pipe' }); }
+
+describe('#2849 — large sync defers extract and queues a stale sweep', () => {
+  beforeAll(async () => {
+    engine = new PGLiteEngine();
+    await engine.connect({});
+    await engine.initSchema();
+    worker = new MinionWorker(engine, { queue: 'test' });
+    await registerBuiltinHandlers(worker, engine, { quiet: true });
+
+    repoPath = mkdtempSync(join(tmpdir(), 'gbrain-large-defer-'));
+    git('git init');
+    git('git config user.email "t@t.com"');
+    git('git config user.name "T"');
+    mkdirSync(join(repoPath, 'people'), { recursive: true });
+    mkdirSync(join(repoPath, 'notes'), { recursive: true });
+    writeFileSync(join(repoPath, 'people/alice.md'), [
+      '---', 'type: person', 'title: Alice', '---', '', 'Alice is a founder.',
+    ].join('\n'));
+    git('git add -A && git commit -m "initial"');
+
+    // Seed: full first sync imports the anchor page + sets last_commit.
+    const { performSync } = await import('../src/commands/sync.ts');
+    await performSync(engine, { repoPath, full: true, noPull: true, noEmbed: true });
+
+    // Second commit: 101 new pages → incremental totalChanges > 100.
+    for (let i = 0; i < 101; i++) {
+      writeFileSync(join(repoPath, `notes/n${i}.md`), [
+        '---', 'type: note', `title: Note ${i}`, '---', '',
+        `[Alice](people/alice) appears in note ${i}.`,
+      ].join('\n'));
+    }
+    git('git add -A && git commit -m "add 101 pages"');
+  }, 120_000);
+
+  afterAll(async () => {
+    if (repoPath) rmSync(repoPath, { recursive: true, force: true });
+    if (engine) await engine.disconnect();
+  }, 60_000);
+
+  test('sync handler defers inline extract and queues extract{stale} follow-up; stale sweep recovers', async () => {
+    const syncHandler = (worker as unknown as { handlers: Map<string, (job: unknown) => Promise<unknown>> })
+      .handlers.get('sync');
+    expect(syncHandler).toBeDefined();
+
+    // Same payload shape the webhook submits (minus embed backfill noise).
+    const result = await syncHandler!({
+      data: { repoPath, noExtract: false, noPull: true, auto_embed_backfill: false },
+      signal: { aborted: false },
+      updateProgress: async () => {},
+    }) as { status: string; extractDeferred?: boolean; extract_stale_job_id?: number | null };
+
+    expect(result.status).toBe('synced');
+    // (a) inline extract was deferred, pages left stale.
+    expect(result.extractDeferred).toBe(true);
+    const staleBefore = await engine.countStalePagesForExtraction();
+    expect(staleBefore).toBeGreaterThan(100);
+    expect(await engine.getLinks('notes/n0')).toHaveLength(0);
+
+    // (b) a follow-up extract job with stale:true was queued.
+    expect(result.extract_stale_job_id).toBeGreaterThan(0);
+    const queue = new MinionQueue(engine);
+    const extractJobs = await queue.getJobs({ name: 'extract', limit: 5 });
+    expect(extractJobs.length).toBe(1);
+    expect((extractJobs[0].data as { stale: boolean }).stale).toBe(true);
+
+    // (c) running the extract handler's stale mode recovers: links + stamps.
+    const extractHandler = (worker as unknown as { handlers: Map<string, (job: unknown) => Promise<unknown>> })
+      .handlers.get('extract');
+    await extractHandler!({
+      data: extractJobs[0].data,
+      signal: { aborted: false },
+      updateProgress: async () => {},
+    });
+    const links = await engine.getLinks('notes/n0');
+    expect(links.some(l => l.to_slug === 'people/alice')).toBe(true);
+    const rows = await engine.executeRaw<{ links_extracted_at: string | null }>(
+      `SELECT links_extracted_at FROM pages WHERE slug = 'notes/n0'`,
+    );
+    expect(rows[0]?.links_extracted_at).not.toBeNull();
+  }, 180_000);
+
+  test('sub-threshold sync does NOT set extractDeferred (no spurious follow-up)', async () => {
+    // One more small commit → inline extract path, no deferral.
+    writeFileSync(join(repoPath, 'notes/small.md'), [
+      '---', 'type: note', 'title: Small', '---', '', 'No big deal.',
+    ].join('\n'));
+    git('git add -A && git commit -m "one small page"');
+    const { performSync } = await import('../src/commands/sync.ts');
+    const result = await performSync(engine, { repoPath, noPull: true, noEmbed: true });
+    expect(result.status).toBe('synced');
+    expect(result.extractDeferred).toBeFalsy();
+  }, 60_000);
+});

test/sync-trigger-cli.test.ts

@@ -100,6 +100,10 @@ describe('runSyncTrigger', () => {
     const job = jobs[0];
     expect(job.priority).toBe(-10);
     expect((job.data as { sourceId: string }).sourceId).toBe('default');
+    // #2849: opt in to inline extraction — the standalone sync handler
+    // defaults noExtract to TRUE, which would leave triggered syncs
+    // extraction-stale.
+    expect((job.data as { noExtract: boolean }).noExtract).toBe(false);
     expect((job.data as { auto_embed_backfill: boolean }).auto_embed_backfill).toBe(true);
   });