test(supervisor): update runner fixture mock for #2750 lock path

tryAcquireDbLock now acquires via engine.executeRaw (signal-boundable) and releases via engine.executeRawDirect instead of the sql tagged template. The fixture's mock engine returned [] from executeRaw, so every spawned supervisor failed lock acquisition and exited LOCK_HELD (2), breaking all 8 supervisor integration tests. The mock now returns the lock row for gbrain_cycle_locks queries and stubs executeRawDirect. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
fix(cycle): enforce the extract-atoms drain deadline end to end (#2750 )
2026-08-17 02:12:40 +00:00 · 2026-07-22 10:53:10 -07:00 · 2026-07-21 14:41:34 -07:00
23 changed files with 685 additions and 650 deletions
@@ -1,7 +1,6 @@
 import type { BrainEngine } from '../core/engine.ts';
 import { embedBatch, currentEmbeddingSignature } from '../core/embedding.ts';
-import type { ChunkInput, ResolvedColumn } from '../core/types.ts';
-import { resolveWriteColumnForEngine } from '../core/search/embedding-column.ts';
+import type { ChunkInput } from '../core/types.ts';
 import { chunkText } from '../core/chunkers/recursive.ts';
 import { createProgress, type ProgressReporter } from '../core/progress.ts';
 import { getCliOptions, cliOptsToProgressOptions } from '../core/cli-options.ts';
@@ -184,13 +183,8 @@ export class EmbeddingDimMismatchError extends Error {
 * fresh-install bug class at the very first invocation instead of letting
 * the worker pool hammer N pages with raw 22000 errors.
 */
-async function preflightDimMismatch(engine: BrainEngine, dryRun: boolean, embeddingColumn?: ResolvedColumn): Promise<void> {
+async function preflightDimMismatch(engine: BrainEngine, dryRun: boolean): Promise<void> {
  if (dryRun) return; // dry-run never embeds, no risk
-  // #1262: an alt-column brain writes to `embeddingColumn`, not the legacy
-  // `embedding` column — the legacy column's dims are irrelevant, and the
-  // registry entry (validated at resolve time) pins the target's dims. Only
-  // the legacy default path needs the schema-vs-gateway dim comparison.
-  if (embeddingColumn && embeddingColumn.name !== 'embedding') return;
  const { readContentChunksEmbeddingDim, embeddingMismatchMessage } = await import('../core/embedding-dim-check.ts');
  const { getEmbeddingDimensions, getEmbeddingModel } = await import('../core/ai/gateway.ts');
  let existing;
@@ -244,12 +238,7 @@ export async function runEmbedCore(engine: BrainEngine, opts: EmbedOpts): Promis
  // v0.37.11.0 (Lane D.2): pre-flight dim-mismatch check. Catches the headline
  // fresh-install bug class before the worker pool spends 20 parallel calls
  // hitting raw Postgres dimension errors.
-  // #1262: resolve the write-side embedding column ONCE at the boundary
-  // (merged config + gateway model) and thread the descriptor through every
-  // upsertChunks / stale-scan below. undefined => legacy `embedding` column.
-  const embeddingColumn = await resolveWriteColumnForEngine(engine);
-
-  await preflightDimMismatch(engine, !!opts.dryRun, embeddingColumn);
+  await preflightDimMismatch(engine, !!opts.dryRun);

  const result: EmbedResult = {
    embedded: 0,
@@ -264,7 +253,7 @@ export async function runEmbedCore(engine: BrainEngine, opts: EmbedOpts): Promis
    for (const s of opts.slugs) {
      if (isAborted(opts.signal)) break; // #1737: stop the per-slug loop on abort
      try {
-        await embedPage(engine, s, !!opts.dryRun, result, opts.sourceId, opts.signal, embeddingColumn);
+        await embedPage(engine, s, !!opts.dryRun, result, opts.sourceId, opts.signal);
      } catch (e: unknown) {
        serr(`  Error embedding ${s}: ${e instanceof Error ? e.message : e}`);
      }
@@ -358,7 +347,7 @@ export async function runEmbedCore(engine: BrainEngine, opts: EmbedOpts): Promis
        catchUp: opts.catchUp,
        pacer,
        paceMaxConcurrency,
-      }, opts.signal, embeddingColumn);
+      }, opts.signal);
    } finally {
      // E1: surface pacing telemetry (human + structured) when pacing was on.
      const snap = pacer.snapshot();
@@ -387,7 +376,7 @@ export async function runEmbedCore(engine: BrainEngine, opts: EmbedOpts): Promis
    return result;
  }
  if (opts.slug) {
-    await embedPage(engine, opts.slug, !!opts.dryRun, result, opts.sourceId, opts.signal, embeddingColumn);
+    await embedPage(engine, opts.slug, !!opts.dryRun, result, opts.sourceId, opts.signal);
    return result;
  }
  throw new Error('No embed target specified. Pass { slug }, { slugs }, { all }, or { stale }.');
@@ -532,13 +521,8 @@ async function embedPage(
  result: EmbedResult,
  sourceId?: string,
  signal?: AbortSignal,
-  embeddingColumn?: ResolvedColumn,
 ) {
  const opts = sourceId ? { sourceId } : undefined;
-  // #1262: write-side descriptor rides only on WRITE calls (upsertChunks).
-  const chunkOpts = (sourceId || embeddingColumn)
-    ? { ...(sourceId && { sourceId }), ...(embeddingColumn && { embeddingColumn }) }
-    : undefined;
  const page = await engine.getPage(slug, opts);
  if (!page) {
    throw new Error(`Page not found: ${slug}`);
@@ -570,7 +554,7 @@ async function embedPage(
    }

    if (inputs.length > 0) {
-      await engine.upsertChunks(slug, inputs, chunkOpts);
+      await engine.upsertChunks(slug, inputs, opts);
      chunks = await engine.getChunks(slug, opts);
    }
  }
@@ -605,7 +589,7 @@ async function embedPage(
    token_count: c.token_count || Math.ceil(c.chunk_text.length / 4),
  }));

-  await engine.upsertChunks(slug, updated, chunkOpts);
+  await engine.upsertChunks(slug, updated, opts);
  // v0.41.31: stamp provenance so a later model/dims swap is detectable as
  // stale. embedPage is the per-slug path used by `gbrain embed <slug>` AND
  // by `gbrain sync`'s post-import embed step (runEmbedCore({slugs})).
@@ -638,7 +622,6 @@ async function embedAll(
    paceMaxConcurrency?: number;
  },
  signal?: AbortSignal,
-  embeddingColumn?: ResolvedColumn,
 ) {
  // v0.41.31: current embedding provenance signature. Stamped onto pages
  // when their chunks are (re)embedded so a later model/dimension swap is
@@ -661,7 +644,7 @@ async function embedAll(
    // D7: thread sourceId so `gbrain embed --stale --source X` actually scopes.
    // v0.41.18.0 (A13): thread batchSize/priority/catchUp into the stale path.
    // #1737: thread the external abort signal so the cycle embed phase bails.
-    return await embedAllStale(engine, sourceId, dryRun, result, onProgress, staleOpts, signature, signal, embeddingColumn);
+    return await embedAllStale(engine, sourceId, dryRun, result, onProgress, staleOpts, signature, signal);
  }

  // --all path: pacer (no-op when off). E-1: lower the worker count to the
@@ -742,10 +725,7 @@ async function embedAll(
        embedding: embeddingMap.get(c.chunk_index) ?? undefined,
        token_count: c.token_count || Math.ceil(c.chunk_text.length / 4),
      }));
-      await observed(pacer, () => engine.upsertChunks(page.slug, updated, {
-        ...(pageSourceId && { sourceId: pageSourceId }),
-        ...(embeddingColumn && { embeddingColumn }),
-      }));
+      await observed(pacer, () => engine.upsertChunks(page.slug, updated, pageOpts));
      // v0.41.31: stamp embedding provenance so a later model swap is
      // detectable as stale.
      await observed(pacer, () =>
@@ -825,16 +805,10 @@ async function embedAllStale(
  },
  signature?: string,
  externalSignal?: AbortSignal,
-  embeddingColumn?: ResolvedColumn,
 ) {
  // D7: thread sourceId so source-scoped runs only count + visit
  // that source's NULL embeddings.
-  // #1262: the stale predicate follows the write-side column — without it an
-  // alt-column brain would perpetually re-select (and re-pay for) chunks whose
-  // target column is already populated.
-  const sourceOpt = (sourceId || embeddingColumn)
-    ? { ...(sourceId && { sourceId }), ...(embeddingColumn && { embeddingColumn }) }
-    : undefined;
+  const sourceOpt = sourceId ? { sourceId } : undefined;

  // v0.41.31: re-embed pages whose embedding_signature drifted (model/dims
  // swap). dry-run must NOT mutate, so it counts signature-stale via the
@@ -993,7 +967,6 @@ async function embedAllStale(
            afterUpdatedAt,
          }),
          ...(sourceId && { sourceId }),
-          ...(embeddingColumn && { embeddingColumn }),
        }),
      );
      if (batch.length === 0) {
@@ -1046,10 +1019,7 @@ async function embedAllStale(
            embedding: staleIdxToEmbedding.get(c.chunk_index) ?? undefined,
            token_count: c.token_count || Math.ceil(c.chunk_text.length / 4),
          }));
-          await observed(pacer, () => engine.upsertChunks(slug, merged, {
-            sourceId: keySourceId,
-            ...(embeddingColumn && { embeddingColumn }),
-          }));
+          await observed(pacer, () => engine.upsertChunks(slug, merged, { sourceId: keySourceId }));
          // v0.41.31: stamp provenance after the page's chunks are embedded —
          // but only when EVERY chunk was stale (fully re-embedded this pass).
          // A partially-stale page keeps preserved chunks of unknown/old
@@ -1120,7 +1090,7 @@ async function embedAllStale(
  // as a clean run — re-running won't help until the underlying failure is fixed.
  if (staleOpts?.catchUp && !effectiveSignal.aborted && embedFailures > 0) {
    const remaining = await engine.countStaleChunks(
-      signature ? { signature, ...sourceOpt } : sourceOpt,
+      signature ? { signature, ...(sourceId ? { sourceId } : {}) } : (sourceId ? { sourceId } : undefined),
    );
    if (remaining > 0) {
      serr(`\n  [embed] catch-up finished but ${remaining} chunk(s) remain stale after ${embedFailures} embed failure(s). These are not embeddable as-is; re-running won't clear them until the underlying error is resolved.`);
@@ -2059,6 +2059,8 @@ export async function registerBuiltinHandlers(
        sourceId,
        windowSeconds,
        brainDir: repoPath,
+        // #2750: worker cancel/timeout/lock-loss propagates into the drain.
+        abortSignal: job.signal,
      });
    } catch (e) {
      if (e instanceof LockUnavailableError) {
@@ -576,16 +576,9 @@ async function runInlineCostGate(

  // Stale backlog: cheap single SQL; fail-open to 0 so a transient DB hiccup
  // never blocks the sync. Signature-aware (model/dims swap surfaces here).
-  // #1262: follow the write-side embedding column — otherwise an alt-column
-  // brain's fully-embedded corpus counts as phantom backlog on every gate.
  let staleChars = 0;
  try {
-    const { resolveWriteColumnForEngine } = await import('../core/search/embedding-column.ts');
-    const embeddingColumn = await resolveWriteColumnForEngine(engine);
-    staleChars = await engine.sumStaleChunkChars({
-      signature: currentEmbeddingSignature(),
-      ...(embeddingColumn && { embeddingColumn }),
-    });
+    staleChars = await engine.sumStaleChunkChars({ signature: currentEmbeddingSignature() });
  } catch {
    staleChars = 0;
  }
@@ -61,7 +61,6 @@ import {
  type SynopsisFailureKind,
 } from './audit-synopsis.ts';
 import type { BrainEngine } from './engine.ts';
-import { resolveWriteColumnForEngine } from './search/embedding-column.ts';
 import type { ChunkInput, CRMode, Page } from './types.ts';
 import type { SourceRow } from './sources-ops.ts';

@@ -287,13 +286,9 @@ export async function reembedPageWithContextualRetrieval(

      // ── PHASE 2: single DB transaction ───────────────────────────
      try {
-        // #1262: contextual re-embeds write TEXT embeddings — thread the
-        // caller-resolved write column like every other embed path.
-        const embeddingColumn = await resolveWriteColumnForEngine(args.engine);
        await args.engine.transaction(async (tx) => {
          await tx.upsertChunks(args.pageSlug, phase1.embeddedChunks, {
            sourceId: args.sourceId,
-            ...(embeddingColumn && { embeddingColumn }),
          });
          await tx.updatePageContextualRetrievalState(
            args.pageSlug,
@@ -23,6 +23,10 @@
 */

 import type { BrainEngine } from '../engine.ts';
+import { anySignal } from '../abort-check.ts';
+
+/** Fresh cleanup budget for the lock release after the window signal fires. */
+const LOCK_RELEASE_GRACE_MS = 5_000;

 export interface ExtractAtomsDrainDeps {
  /**
@@ -30,13 +34,13 @@ export interface ExtractAtomsDrainDeps {
   * via `withRefreshingLock`. MUST throw when the lock is held by another
   * process (e.g. `LockUnavailableError`) — the drain lets that propagate so
   * the caller can report `cycle_already_running` and exit, matching the
-   * routine cycle's skip contract.
+   * routine cycle's skip contract. The signal bounds lock acquisition too.
   */
-  withLock: <T>(work: () => Promise<T>) => Promise<T>;
-  /** Process one bounded batch (rediscovers eligibility). Returns counts. */
-  runBatch: () => Promise<{ extracted: number; skipped: number }>;
+  withLock: <T>(work: () => Promise<T>, signal: AbortSignal) => Promise<T>;
+  /** Process one batch. The signal fires at the drain wallclock deadline. */
+  runBatch: (signal: AbortSignal) => Promise<{ extracted: number; skipped: number }>;
  /** Count remaining eligible-but-unextracted pages, or null on query error. */
-  countRemaining: () => Promise<number | null>;
+  countRemaining: (signal: AbortSignal) => Promise<number | null>;
  /** Injectable clock. Production: Date.now. */
  now: () => number;
  /** Optional progress sink (one line per batch). */
@@ -48,6 +52,8 @@ export interface ExtractAtomsDrainOpts {
  windowMs: number;
  /** Hard cap on batches (belt-and-suspenders against a 0-progress loop). Default 1000. */
  maxBatches?: number;
+  /** External caller cancellation (worker timeout / shutdown). */
+  abortSignal?: AbortSignal;
 }

 export interface ExtractAtomsDrainResult {
@@ -68,35 +74,79 @@ export async function runExtractAtomsDrain(
  opts: ExtractAtomsDrainOpts,
 ): Promise<ExtractAtomsDrainResult> {
  const maxBatches = opts.maxBatches ?? 1000;
-  return deps.withLock(async () => {
-    const deadline = deps.now() + opts.windowMs;
+  const deadline = deps.now() + opts.windowMs;
+  // #2750: the window used to be checked only BETWEEN batches, so one slow
+  // batch (sequential LLM calls) or a hung lock/count/write overran it without
+  // bound (observed window=120s → 282.5s). A real-time deadline signal now
+  // cancels (Postgres) or abandons (PGLite, cooperative) whatever is in
+  // flight; the injected clock still drives loop-boundary checks so the pure
+  // loop stays unit-testable.
+  const signal = anySignal(
+    AbortSignal.timeout(Math.max(1, opts.windowMs)),
+    opts.abortSignal,
+  );
+  const result: ExtractAtomsDrainResult = await deps.withLock(async () => {
    let extracted = 0;
    let skipped = 0;
    let batches = 0;
    let stopped: ExtractAtomsDrainResult['stopped'] = 'window';

-    while (deps.now() < deadline) {
+    while (deps.now() < deadline && !signal.aborted) {
      if (batches >= maxBatches) { stopped = 'max_batches'; break; }

-      const before = await deps.countRemaining();
+      let before: number | null;
+      try {
+        before = await deps.countRemaining(signal);
+      } catch (err) {
+        if (signal.aborted) break;
+        throw err;
+      }
      if (before === 0) { stopped = 'drained'; break; }

-      const r = await deps.runBatch();
+      // The backlog count consumed the same wallclock budget — re-check so a
+      // slow count can't hand the batch a window that already expired.
+      if (deps.now() >= deadline || signal.aborted) break;
+
+      let r: { extracted: number; skipped: number };
+      try {
+        r = await deps.runBatch(signal);
+      } catch (err) {
+        if (signal.aborted) break;
+        throw err;
+      }
      extracted += r.extracted;
      skipped += r.skipped;
      batches++;
      deps.onBatch?.({ batch: batches, extracted: r.extracted, remaining: before });

+      // A deadline abort inside the batch can surface as zero progress;
+      // window exhaustion wins over the generic no_progress label.
+      if (deps.now() >= deadline || signal.aborted) break;
+
      // Stop if a batch made zero forward progress — extraction is failing or
      // everything left is ineligible (e.g. all skipped). Prevents a hot loop
      // that spends budget without draining.
      if (r.extracted === 0 && r.skipped === 0) { stopped = 'no_progress'; break; }
    }

-    const remaining = await deps.countRemaining();
+    // After the window elapsed, don't spend more unbounded time on a final
+    // count — report remaining as unknown instead of overrunning further.
+    const windowElapsed = signal.aborted || deps.now() >= deadline;
+    let remaining: number | null = null;
+    if (!windowElapsed) {
+      try {
+        remaining = await deps.countRemaining(signal);
+      } catch (err) {
+        if (!signal.aborted) throw err;
+      }
+    }
    if (remaining === 0) stopped = 'drained';
    return { phase: 'extract_atoms', status: 'ok', extracted, skipped, remaining, batches, stopped };
-  });
+  }, signal);
+  // Internal window expiry is a normal partial result. An EXTERNAL abort
+  // (worker cancel/timeout/shutdown) must reject so Minion records the abort.
+  if (opts.abortSignal?.aborted) throw opts.abortSignal.reason;
+  return result;
 }

 // ─── Shared wiring helper (v0.42.x #1685 DECISION 5A) ──────────────────────
@@ -134,6 +184,8 @@ export interface DrainForSourceOpts {
  maxBatches?: number;
  /** Optional per-batch progress sink (stderr line in dream; job progress in the handler). */
  onBatch?: ExtractAtomsDrainDeps['onBatch'];
+  /** Worker cancellation / shutdown signal (Minion `job.signal`). */
+  abortSignal?: AbortSignal;
 }

 export async function runExtractAtomsDrainForSource(
@@ -149,12 +201,17 @@ export async function runExtractAtomsDrainForSource(

  return runExtractAtomsDrain(
    {
-      withLock: (work) => withRefreshingLock(engine, lockId, work, { ttlMinutes: 5 }),
-      runBatch: async () => {
+      withLock: (work, signal) => withRefreshingLock(engine, lockId, work, {
+        ttlMinutes: 5,
+        signal,
+        releaseTimeoutMs: LOCK_RELEASE_GRACE_MS,
+      }),
+      runBatch: async (signal) => {
        const r = await runPhaseExtractAtoms(engine, {
          sourceId: extractionSourceId,
          dryRun: false,
          brainDir: opts.brainDir,
+          abortSignal: signal,
        });
        const d = (r.details ?? {}) as Record<string, unknown>;
        return {
@@ -162,10 +219,14 @@ export async function runExtractAtomsDrainForSource(
          skipped: Number(d.duplicates_skipped ?? 0),
        };
      },
-      countRemaining: () => countExtractAtomsBacklog(engine, extractionSourceId),
+      countRemaining: (signal) => countExtractAtomsBacklog(engine, extractionSourceId, signal),
      now: Date.now,
      onBatch: opts.onBatch,
    },
-    { windowMs: opts.windowSeconds * 1000, maxBatches: opts.maxBatches },
+    {
+      windowMs: opts.windowSeconds * 1000,
+      maxBatches: opts.maxBatches,
+      abortSignal: opts.abortSignal,
+    },
  );
 }
@@ -58,6 +58,10 @@ import { createHash } from 'crypto';
 import { slugifySegment } from '../sync.ts';

 const DEFAULT_BUDGET_USD = 0.3;
+// #2750: fresh wallclock budget for the receipt/rollup bookkeeping writes when
+// the caller's deadline already fired — committed atoms must not lose their
+// cost/receipt trail, but the writes can't be unbounded either.
+const BOOKKEEPING_GRACE_MS = 5_000;

 // v0.42+ TODO: read atom_type enum from active pack manifest at runtime.
 const ATOM_TYPES = [
@@ -155,6 +159,13 @@ export interface ExtractAtomsOpts {
   * `heartbeat()` on the passed reporter.
   */
  progress?: ProgressReporter;
+  /**
+   * #2750: caller deadline/cancellation. Forwarded to every gateway call and
+   * DB query/write so the drain window bounds real lifetime, plus a
+   * cooperative between-item check (the PGLite path, where query abort only
+   * abandons the waiter).
+   */
+  abortSignal?: AbortSignal;
 }

 interface ExtractedAtom {
@@ -212,6 +223,7 @@ export async function discoverExtractablePages(
  engine: BrainEngine,
  sourceId: string,
  affectedSlugs?: string[],
+  abortSignal?: AbortSignal,
 ): Promise<DiscoveredPage[]> {
  const hasFilter = Array.isArray(affectedSlugs) && affectedSlugs.length > 0;
  const sql = `
@@ -251,13 +263,16 @@ export async function discoverExtractablePages(
      slug: string;
      compiled_truth: string;
      content_hash: string;
-    }>(sql, params);
+    }>(sql, params, { signal: abortSignal });
    return rows.map((r) => ({
      slug: r.slug,
      content: r.compiled_truth,
      contentHash: r.content_hash,
    }));
  } catch (err) {
+    // A deadline abort is not a fail-soft condition — propagate so the
+    // caller stops instead of proceeding with an empty page list.
+    if (abortSignal?.aborted) throw err;
    const msg = err instanceof Error ? err.message : String(err);
    console.error(`[extract_atoms] page-discovery query failed: ${msg}`);
    return []; // fail-soft: transcript path still proceeds
@@ -282,6 +297,7 @@ export async function discoverExtractablePages(
 export async function countExtractAtomsBacklog(
  engine: BrainEngine,
  sourceId?: string,
+  abortSignal?: AbortSignal,
 ): Promise<number | null> {
  try {
    // Two modes: scoped (the phase's per-source `remaining`) vs brain-wide
@@ -321,9 +337,10 @@ export async function countExtractAtomsBacklog(
    const params = scoped
      ? [sourceId, extractableTypes, MIN_PAGE_CHARS_FOR_EXTRACTION]
      : [extractableTypes, MIN_PAGE_CHARS_FOR_EXTRACTION];
-    const rows = await engine.executeRaw<{ cnt: string | number }>(sql, params);
+    const rows = await engine.executeRaw<{ cnt: string | number }>(sql, params, { signal: abortSignal });
    return Number(rows[0]?.cnt ?? 0);
  } catch (err) {
+    if (abortSignal?.aborted) throw err;
    const msg = err instanceof Error ? err.message : String(err);
    console.error(`[extract_atoms] backlog count failed: ${msg}`);
    return null;
@@ -350,6 +367,7 @@ export async function atomsExistingForHashes(
  engine: BrainEngine,
  sourceId: string,
  contentHash16s: string[],
+  abortSignal?: AbortSignal,
 ): Promise<Set<string>> {
  if (contentHash16s.length === 0) return new Set();
  try {
@@ -361,9 +379,11 @@ export async function atomsExistingForHashes(
          AND deleted_at IS NULL
          AND frontmatter->>'source_hash' = ANY($2::text[])`,
      [sourceId, contentHash16s],
+      { signal: abortSignal },
    );
    return new Set(rows.map(r => r.h));
  } catch (err) {
+    if (abortSignal?.aborted) throw err;
    const msg = err instanceof Error ? err.message : String(err);
    console.error(`[extract_atoms] batch idempotency check failed (assuming none extracted): ${msg}`);
    return new Set();
@@ -384,6 +404,7 @@ export async function runPhaseExtractAtoms(
 ): Promise<PhaseResult> {
  const sourceId = opts.sourceId ?? 'default';
  const chat = opts._chat ?? gatewayChat;
+  if (opts.abortSignal?.aborted) throw opts.abortSignal.reason;

  // 1a. Get transcripts (test seam OR production discovery).
  //     v0.41.2.1: config loader switched to loadConfigWithEngine() so the
@@ -425,7 +446,7 @@ export async function runPhaseExtractAtoms(
  if (opts._pages !== undefined) {
    pages = opts._pages;
  } else {
-    pages = await discoverExtractablePages(engine, sourceId, opts.affectedSlugs);
+    pages = await discoverExtractablePages(engine, sourceId, opts.affectedSlugs, opts.abortSignal);
  }

  // 2. Apply transcript-side source-hash idempotency in ONE batch query
@@ -437,7 +458,7 @@ export async function runPhaseExtractAtoms(
  // Surface a heartbeat before the batch query so even an instant
  // short-circuit shows a sign of life (closes Issue 2 silent-phase pain).
  opts.progress?.heartbeat(`checking existing atoms for ${allHashes16.length} transcripts`);
-  const existingHashes = await atomsExistingForHashes(engine, sourceId, allHashes16);
+  const existingHashes = await atomsExistingForHashes(engine, sourceId, allHashes16, opts.abortSignal);
  for (const t of transcripts) {
    if (existingHashes.has(t.contentHash.slice(0, 16))) {
      duplicatesSkipped++;
@@ -501,6 +522,7 @@ export async function runPhaseExtractAtoms(
  const failures: Array<{ source: string; error: string }> = [];
  let estimatedSpendUsd = 0;
  const budgetCap = DEFAULT_BUDGET_USD;
+  let deadlineAborted = false;

  // v0.41.19.0 (T3): throttled yield helper. Fires `opts.yieldDuringPhase`
  // every 30s. Cycle.ts threads `buildYieldDuringPhase(lock, outer)` so
@@ -526,6 +548,12 @@ export async function runPhaseExtractAtoms(
  }

  for (const item of work) {
+    // #2750: cooperative between-item abort. Works on every engine — this is
+    // the primary bound on PGLite, where query abort only abandons the waiter.
+    if (opts.abortSignal?.aborted) {
+      deadlineAborted = true;
+      break;
+    }
    await maybeYield();
    if (estimatedSpendUsd >= budgetCap) {
      if (item.kind === 'transcript') transcriptsSkipped++;
@@ -544,16 +572,22 @@ export async function runPhaseExtractAtoms(
          },
        ],
        maxTokens: 2000,
+        abortSignal: opts.abortSignal,
      });
+      // Rough cost estimate — Haiku at ~$0.80/M input + $4/M output.
+      // A completed gateway call is billable even if the deadline fires
+      // immediately afterward, so record usage BEFORE the abort check.
+      estimatedSpendUsd +=
+        (result.usage.input_tokens * 0.8 + result.usage.output_tokens * 4.0) / 1_000_000;
+      if (opts.abortSignal?.aborted) {
+        deadlineAborted = true;
+        break;
+      }
      // Post-await yield: closes the "long LLM call past TTL" hazard
      // codex flagged. The 30s throttle inside maybeYield bounds the
      // actual refresh rate so this is cheap when calls are fast.
      await maybeYield();

-      // Rough cost estimate — Haiku at ~$0.80/M input + $4/M output
-      estimatedSpendUsd +=
-        (result.usage.input_tokens * 0.8 + result.usage.output_tokens * 4.0) / 1_000_000;
-
      const atoms = parseAtomsResponse(result.text);
      if (atoms.length === 0) {
        if (item.kind === 'transcript') transcriptsProcessed++;
@@ -592,7 +626,7 @@ export async function runPhaseExtractAtoms(
              },
              timeline: '',
            },
-            { sourceId },
+            { sourceId, signal: opts.abortSignal },
          );
          totalAtomsExtracted++;
        }
@@ -605,6 +639,11 @@ export async function runPhaseExtractAtoms(
      // Reporter rate-limits to ~1 line/sec; safe to tick every iter.
      opts.progress?.tick(1, `${totalAtomsExtracted} atoms / ${duplicatesSkipped} skipped`);
    } catch (err) {
+      // A deadline abort is a partial result, not a per-item failure.
+      if (opts.abortSignal?.aborted) {
+        deadlineAborted = true;
+        break;
+      }
      failures.push({
        source: originLabel,
        error: err instanceof Error ? err.message : String(err),
@@ -615,6 +654,12 @@ export async function runPhaseExtractAtoms(
  // v0.42 Wave B2: write extract receipt + rollup row when the phase
  // actually extracted atoms. Both are best-effort per F-OUT-19 —
  // audit-trail / search-visibility surfaces don't block the phase result.
+  //
+  // #2750: bookkeeping runs on a FRESH short grace signal, never the caller's
+  // work deadline — the deadline may have already fired (partial run) and
+  // committed atoms must not lose their receipt/cost trail; but the writes
+  // stay bounded so the overrun is capped at the grace window.
+  const bookkeepingSignal = opts.dryRun ? undefined : AbortSignal.timeout(BOOKKEEPING_GRACE_MS);
  if (!opts.dryRun && totalAtomsExtracted > 0) {
    const runId = `atoms-${Date.now().toString(36)}-${sourceId.slice(0, 4)}`;
    try {
@@ -629,19 +674,24 @@ export async function runPhaseExtractAtoms(
        summary:
          `Extracted ${totalAtomsExtracted} atoms from ` +
          `${transcriptsProcessed} transcripts + ${pagesProcessed} pages.`,
-      });
+      }, { signal: bookkeepingSignal });
    } catch (err) {
      console.error(`[extract_atoms] receipt write failed: ${(err as Error).message}`);
    }
  }
  if (!opts.dryRun) {
-    await upsertExtractRollup(engine, {
-      kind: 'atoms',
-      source_id: sourceId,
-      cost_delta: estimatedSpendUsd,
-      round_completed_delta: failures.length === 0 ? 1 : 0,
-      halt_delta: failures.length > 0 ? 1 : 0,
-    });
+    try {
+      await upsertExtractRollup(engine, {
+        kind: 'atoms',
+        source_id: sourceId,
+        cost_delta: estimatedSpendUsd,
+        // A deadline-truncated run is not a completed round.
+        round_completed_delta: failures.length === 0 && !deadlineAborted ? 1 : 0,
+        halt_delta: failures.length > 0 ? 1 : 0,
+      }, { signal: bookkeepingSignal });
+    } catch (err) {
+      console.error(`[extract_atoms] rollup write failed: ${(err as Error).message}`);
+    }
  }

  return {
@@ -670,6 +720,7 @@ export async function runPhaseExtractAtoms(
      budget_usd: budgetCap,
      source_id: sourceId,
      dry_run: opts.dryRun ?? false,
+      deadline_aborted: deadlineAborted,
    },
  };
 }
@@ -26,7 +26,8 @@ import type { BrainEngine } from './engine.ts';

 export interface DbLockHandle {
  id: string;
-  release: () => Promise<void>;
+  /** Optional signal bounds the release DELETE (deadline-bound callers). */
+  release: (signal?: AbortSignal) => Promise<void>;
  refresh: () => Promise<void>;
 }

@@ -173,6 +174,7 @@ export async function tryAcquireDbLock(
  engine: BrainEngine,
  lockId: string,
  ttlMinutes: number = DEFAULT_TTL_MINUTES,
+  opts: { signal?: AbortSignal } = {},
 ): Promise<DbLockHandle | null> {
  const pid = process.pid;
  const host = hostname();
@@ -205,20 +207,26 @@ export async function tryAcquireDbLock(
    // `gbrain sync --break-lock --max-age <s>` uses last_refreshed_at (not
    // acquired_at) to identify wedged-but-alive holders without stealing
    // healthy long-running holders that are actively refreshing.
-    const rows: Array<{ id: string }> = await sql`
-      INSERT INTO gbrain_cycle_locks (id, holder_pid, holder_host, acquired_at, ttl_expires_at, last_refreshed_at)
-      VALUES (${lockId}, ${pid}, ${host}, NOW(), NOW() + ${ttl}::interval, NOW())
-      ON CONFLICT (id) DO UPDATE
-        SET holder_pid = ${pid},
-            holder_host = ${host},
-            acquired_at = NOW(),
-            ttl_expires_at = NOW() + ${ttl}::interval,
-            last_refreshed_at = NOW()
-        WHERE gbrain_cycle_locks.ttl_expires_at < NOW()
-          AND (gbrain_cycle_locks.last_refreshed_at IS NULL
-               OR gbrain_cycle_locks.last_refreshed_at < NOW() - ${stealGraceSeconds} * INTERVAL '1 second')
-      RETURNING id
-    `;
+    // #2750: routed through executeRaw so a deadline-bound caller's signal
+    // can cancel a hung acquire (pool exhaustion). Cancellation is
+    // transactional; in the rare ambiguous-commit case the row's TTL is the
+    // backstop (drain locks use a short 5-minute TTL).
+    const rows = await engine.executeRaw<{ id: string }>(
+      `INSERT INTO gbrain_cycle_locks (id, holder_pid, holder_host, acquired_at, ttl_expires_at, last_refreshed_at)
+       VALUES ($1, $2, $3, NOW(), NOW() + $4::interval, NOW())
+       ON CONFLICT (id) DO UPDATE
+         SET holder_pid = $2,
+             holder_host = $3,
+             acquired_at = NOW(),
+             ttl_expires_at = NOW() + $4::interval,
+             last_refreshed_at = NOW()
+         WHERE gbrain_cycle_locks.ttl_expires_at < NOW()
+           AND (gbrain_cycle_locks.last_refreshed_at IS NULL
+                OR gbrain_cycle_locks.last_refreshed_at < NOW() - $5 * INTERVAL '1 second')
+       RETURNING id`,
+      [lockId, pid, host, ttl, stealGraceSeconds],
+      { signal: opts.signal },
+    );
    if (rows.length === 0) return null;
    const deregister = registerCleanup(`db-lock:${lockId}`, async () => {
      await sql`
@@ -241,12 +249,17 @@ export async function tryAcquireDbLock(
          [ttl, lockId, pid],
        );
      },
-      release: async () => {
+      release: async (signal?: AbortSignal) => {
        deregister();
-        await sql`
-          DELETE FROM gbrain_cycle_locks
-          WHERE id = ${lockId} AND holder_pid = ${pid}
-        `;
+        // Direct session pool (same rationale as refresh, #1794) + optional
+        // signal so a deadline-bound caller's release can't hang forever on
+        // an exhausted pooler. TTL is the backstop if the DELETE is cancelled.
+        await engine.executeRawDirect(
+          `DELETE FROM gbrain_cycle_locks
+            WHERE id = $1 AND holder_pid = $2`,
+          [lockId, pid],
+          { signal },
+        );
      },
    };
  }
@@ -303,6 +316,11 @@ export async function tryAcquireDbLock(
  const first = await acquireOnce();
  if (first) return first;

+  // #2750: deadline-bound callers prefer an honest busy result over the
+  // best-effort same-host takeover below, whose inspect/delete/retry calls
+  // are not signal-bounded. The initial upsert already reclaims expired locks.
+  if (opts.signal) return null;
+
  // v0.42 (#1780 Gap 3): the lock is held and its TTL hasn't expired (the
  // upsert's ON CONFLICT ... WHERE ttl_expires_at < NOW() returned no row).
  // If the holder is on THIS host, provably dead, and past the grace window,
@@ -796,6 +814,10 @@ export interface WithRefreshingLockOpts {
  ttlMinutes?: number;
  /** Heartbeat-fail threshold in ms — abort if SELECT 1 takes longer. Default 30000. */
  heartbeatTimeoutMs?: number;
+  /** #2750: bound lock acquisition with the caller's deadline signal. */
+  signal?: AbortSignal;
+  /** Fresh cleanup budget for the release DELETE when `signal` is set. Default 5000. */
+  releaseTimeoutMs?: number;
 }

 /**
@@ -815,7 +837,7 @@ export async function withRefreshingLock<T>(
  // Refresh 6x per TTL window so a missed tick doesn't expire the lock.
  const refreshIntervalMs = Math.max(15000, (ttlMinutes * 60 * 1000) / 6);

-  const handle = await tryAcquireDbLock(engine, lockId, ttlMinutes);
+  const handle = await tryAcquireDbLock(engine, lockId, ttlMinutes, { signal: opts.signal });
  if (!handle) throw new LockUnavailableError(lockId);

  let healthOk = true;
@@ -854,7 +876,13 @@ export async function withRefreshingLock<T>(
    return await work();
  } finally {
    clearInterval(interval);
-    try { await handle.release(); } catch { /* idempotent */ }
+    // #2750: when the caller is deadline-bound, its work signal may already
+    // have fired — release on a FRESH short grace signal so cleanup neither
+    // inherits the spent deadline nor hangs unbounded. TTL is the backstop.
+    const releaseSignal = opts.signal
+      ? AbortSignal.timeout(opts.releaseTimeoutMs ?? 5_000)
+      : undefined;
+    try { await handle.release(releaseSignal); } catch { /* idempotent; TTL backstop */ }
    if (!healthOk) {
      // Surface that the heartbeat detected backend trouble — caller can
      // log to the connection-events audit if desired.
@@ -18,7 +18,7 @@
 */

 import type { BrainEngine } from './engine.ts';
-import type { ChunkInput, ResolvedColumn } from './types.ts';
+import type { ChunkInput } from './types.ts';
 import { embedBatchWithBackoff } from '../commands/embed.ts';
 import { type DbPacer, createNoopPacer, observed } from './db-pacer.ts';
 import { AbortError } from './abort-check.ts';
@@ -61,13 +61,6 @@ export interface EmbedStaleOpts {
   * Omit to keep the legacy `embedding IS NULL`-only behavior.
   */
  embeddingSignature?: string;
-  /**
-   * #1262: caller-resolved write-side embedding column. Threaded into BOTH
-   * listStaleChunks (staleness predicate) and upsertChunks (write target) so
-   * an alt-column brain converges instead of re-selecting embedded rows.
-   * Resolve at the boundary via `resolveWriteColumnForEngine()`.
-   */
-  embeddingColumn?: ResolvedColumn;
  /**
   * DB-contention pacer (paced-backfill). When enabled it (a) supplies the
   * worker count via the caller passing `concurrency = bundle.maxConcurrency`
@@ -163,7 +156,6 @@ export async function embedStaleForSource(
        afterPageId,
        afterChunkIndex,
        sourceId,
-        ...(opts.embeddingColumn && { embeddingColumn: opts.embeddingColumn }),
      }),
    );
    if (batch.length === 0) {
@@ -231,10 +223,7 @@ export async function embedStaleForSource(
          doc_comment: c.doc_comment ?? undefined,
          symbol_name_qualified: c.symbol_name_qualified ?? undefined,
        }));
-        await observed(pacer, () => engine.upsertChunks(slug, merged, {
-          sourceId: keySourceId,
-          ...(opts.embeddingColumn && { embeddingColumn: opts.embeddingColumn }),
-        }));
+        await observed(pacer, () => engine.upsertChunks(slug, merged, { sourceId: keySourceId }));
        // v0.41.31: stamp provenance only when EVERY chunk was stale (fully
        // re-embedded this pass) — a partially-stale page keeps preserved
        // chunks of unknown provenance, so don't claim current. After the
@@ -12,7 +12,6 @@ import type {
  BrainStats, BrainHealth,
  IngestLogEntry, IngestLogInput,
  EngineConfig,
-  ResolvedColumn,
  CodeEdgeInput, CodeEdgeResult,
  EvalCandidate, EvalCandidateInput,
  EvalCaptureFailure, EvalCaptureFailureReason,
@@ -697,8 +696,16 @@ export interface BrainEngine {
   * is included in the INSERT column list so ON CONFLICT (source_id, slug)
   * DO UPDATE actually targets the intended row instead of fabricating a
   * duplicate at (default, slug). Multi-source brains MUST pass sourceId.
+   *
+   * `opts.signal` (#2750): optional cancellation for deadline-bound writers.
+   * Postgres cancels the in-flight statement; PGLite pre-checks only (query
+   * cancellation is not possible in-process — cooperative abort between calls).
   */
-  putPage(slug: string, page: PageInput, opts?: { sourceId?: string }): Promise<Page>;
+  putPage(
+    slug: string,
+    page: PageInput,
+    opts?: { sourceId?: string; signal?: AbortSignal },
+  ): Promise<Page>;
  /**
   * v0.41.13 (#1309) — identity-based dedup pre-check for the import pipeline.
   *
@@ -988,13 +995,8 @@ export interface BrainEngine {
   * — Postgres rolls back automatically on conn drop, so commit-ambiguous
   * failure replays to the same end state. Callers MUST NOT wrap externally;
   * see {@link BatchOpts} retry-contract block.
-   *
-   * `opts.embeddingColumn` (optional) selects the content_chunks column that
-   * receives TEXT embeddings (#1262). The caller resolves the descriptor at
-   * the import/embed boundary via `resolveWriteColumn()`; engines never read
-   * config or choose columns themselves. Omitted => legacy `embedding`.
   */
-  upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string; embeddingColumn?: ResolvedColumn } & BatchOpts): Promise<void>;
+  upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string } & BatchOpts): Promise<void>;
  /**
   * Read every chunk for a page. `opts.sourceId` source-scopes the page
   * lookup; without it, multi-source brains return chunks from every
@@ -1011,13 +1013,8 @@ export interface BrainEngine {
   * counts across every source in the brain. Operators running
   * `gbrain embed --stale --source media-corpus` expect only that
   * source's NULLs touched; the caller threads `sourceId` here.
-   *
-   * `opts.embeddingColumn` switches the staleness predicate from the legacy
-   * `embedding` column to the resolved write-side column, so alt-column
-   * brains do not perpetually re-select rows whose target column is already
-   * populated (#1262). Must match the eventual upsertChunks target.
   */
-  countStaleChunks(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number>;
+  countStaleChunks(opts?: { sourceId?: string; signature?: string }): Promise<number>;
  /**
   * Sum of LENGTH(chunk_text) over stale chunks — the character-count
   * backlog the embed phase / embed-backfill will process. Sibling of
@@ -1031,13 +1028,8 @@ export interface BrainEngine {
   * model signature (a model/dims swap). NULL signature is GRANDFATHERED
   * (never counted) so the post-migration corpus isn't flagged en masse.
   * Omit `signature` for the legacy `embedding IS NULL`-only count.
-   *
-   * `opts.embeddingColumn` switches the staleness predicate to the resolved
-   * write-side column (#1262) — same contract as countStaleChunks — so the
-   * sync cost gate doesn't count an alt-column brain's fully-embedded corpus
-   * as phantom backlog.
   */
-  sumStaleChunkChars(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number>;
+  sumStaleChunkChars(opts?: { sourceId?: string; signature?: string }): Promise<number>;
  /**
   * Stamp `pages.embedding_signature = signature` for one page. Called after
   * a page's chunks are (re)embedded so a later model swap can detect it as
@@ -1085,9 +1077,6 @@ export interface BrainEngine {
    // both round-trip TIMESTAMPTZ as Date | string; ISO string is the
    // common denominator on the wire).
    afterUpdatedAt?: string | null;
-    // #1262: staleness predicate targets this column when set (must match
-    // countStaleChunks and the eventual upsertChunks write target).
-    embeddingColumn?: ResolvedColumn;
  }): Promise<StaleChunkRow[]>;
  /**
   * Delete every chunk for a page. Internal page-id lookup is sourceId-scoped
@@ -187,6 +187,7 @@ function buildReceiptFrontmatter(input: ExtractReceiptInput): Record<string, unk
 export async function writeReceipt(
  engine: BrainEngine,
  input: ExtractReceiptInput,
+  opts?: { signal?: AbortSignal },
 ): Promise<{ slug: string; page: Page }> {
  const slug = receiptSlug(input);
  const title = `${input.kind} — ${input.round} — ${input.source_id}`;
@@ -201,7 +202,7 @@ export async function writeReceipt(
      compiled_truth,
      frontmatter,
    },
-    { sourceId: input.source_id },
+    { sourceId: input.source_id, signal: opts?.signal },
  );

  return { slug, page };
@@ -70,6 +70,7 @@ function today(): string {
 export async function upsertExtractRollup(
  engine: BrainEngine,
  input: RollupUpsertInput,
+  opts?: { signal?: AbortSignal },
 ): Promise<{ ok: boolean; error?: string }> {
  const day = input.day ?? today();
  const cost = input.cost_delta ?? 0;
@@ -96,9 +97,12 @@ export async function upsertExtractRollup(
         rollup_write_failures  = extract_rollup_7d.rollup_write_failures  + EXCLUDED.rollup_write_failures,
         updated_at             = now()`,
      [input.kind, input.source_id, day, cost, halts, evalFails, evalPasses, completed, failures],
+      { signal: opts?.signal },
    );
    return { ok: true };
  } catch (err) {
+    // Signal-bounded callers get the abort surfaced, not a swallowed `ok:false`.
+    if (opts?.signal?.aborted) throw err;
    const msg = (err as Error).message || String(err);
    // Don't spam: log once per process per (kind, day) error class.
    rollupErrorLogOnce(input.kind, day, msg);
@@ -10,8 +10,7 @@ import { findChunkForOffset } from './chunkers/edge-extractor.ts';
 import { extractCodeRefs, imageOfCandidates } from './link-extraction.ts';
 import { embedBatch, embedMultimodal, currentEmbeddingSignature } from './embedding.ts';
 import { slugifyPath, slugifyCodePath, isCodeFilePath } from './sync.ts';
-import type { ChunkInput, PageInput, PageType, ResolvedColumn } from './types.ts';
-import { resolveWriteColumnForEngine } from './search/embedding-column.ts';
+import type { ChunkInput, PageInput, PageType } from './types.ts';
 import { computeEffectiveDate } from './effective-date.ts';
 import { MARKDOWN_CHUNKER_VERSION } from './chunkers/recursive.ts';
 import { logSlugFallback } from './audit-slug-fallback.ts';
@@ -741,14 +740,6 @@ export async function importFromContent(
  // schema DEFAULT — required for multi-source brains; harmless ('default')
  // for single-source callers.
  const txOpts = sourceId ? { sourceId } : undefined;
-  // #1262: resolve the write-side embedding column once (merged config +
-  // gateway model) BEFORE the transaction; the descriptor rides only on
-  // upsertChunks so text embeddings land in the registered column.
-  const chunkWriteColumn = await resolveWriteColumnForEngine(engine);
-  const chunkOpts: { sourceId?: string; embeddingColumn?: ResolvedColumn } | undefined =
-    (sourceId || chunkWriteColumn)
-      ? { ...(sourceId && { sourceId }), ...(chunkWriteColumn && { embeddingColumn: chunkWriteColumn }) }
-      : undefined;
  await engine.transaction(async (tx) => {
    if (existing) await tx.createVersion(slug, txOpts);

@@ -833,7 +824,7 @@ export async function importFromContent(
    }

    if (chunks.length > 0) {
-      await tx.upsertChunks(slug, chunks, chunkOpts);
+      await tx.upsertChunks(slug, chunks, txOpts);
      // v0.41.31: stamp embedding provenance when this import actually
      // embedded (not --no-embed), so a later model/dims swap is detectable
      // as stale via embed --stale. The deferred/backfill + per-slug embed
@@ -1073,12 +1064,6 @@ export async function importCodeFile(
  const title = `${relativePath} (${lang})`;
  const sourceId = opts.sourceId;
  const txOpts = sourceId ? { sourceId } : undefined;
-  // #1262: write-side embedding column descriptor (rides only on upsertChunks).
-  const chunkWriteColumn = await resolveWriteColumnForEngine(engine);
-  const chunkOpts: { sourceId?: string; embeddingColumn?: ResolvedColumn } | undefined =
-    (sourceId || chunkWriteColumn)
-      ? { ...(sourceId && { sourceId }), ...(chunkWriteColumn && { embeddingColumn: chunkWriteColumn }) }
-      : undefined;

  const byteLength = Buffer.byteLength(content, 'utf-8');
  if (byteLength > MAX_FILE_SIZE) {
@@ -1198,7 +1183,7 @@ export async function importCodeFile(
    await tx.addTag(slug, lang, txOpts);

    if (chunks.length > 0) {
-      await tx.upsertChunks(slug, chunks, chunkOpts);
+      await tx.upsertChunks(slug, chunks, txOpts);
      // v0.41.31: stamp embedding provenance ONLY when every chunk was
      // freshly embedded with the current model this call (no reuse-by-hash
      // carrying old-model vectors). Mixed pages stay unstamped rather than
@@ -1347,12 +1332,6 @@ export async function withImportTransaction(
 ): Promise<void> {
  const sourceId = spec.sourceId ?? 'default';
  const txOpts = spec.sourceId ? { sourceId: spec.sourceId } : undefined;
-  // #1262: write-side embedding column descriptor (rides only on upsertChunks).
-  const chunkWriteColumn = await resolveWriteColumnForEngine(engine);
-  const chunkOpts: { sourceId?: string; embeddingColumn?: ResolvedColumn } | undefined =
-    (spec.sourceId || chunkWriteColumn)
-      ? { ...(spec.sourceId && { sourceId: spec.sourceId }), ...(chunkWriteColumn && { embeddingColumn: chunkWriteColumn }) }
-      : undefined;
  await engine.transaction(async (tx) => {
    if (spec.hadExisting) await tx.createVersion(spec.slug, txOpts);
    await tx.putPage(spec.slug, spec.page, txOpts);
@@ -1368,7 +1347,7 @@ export async function withImportTransaction(
    }
    if (spec.chunks !== undefined) {
      if (spec.chunks.length > 0) {
-        await tx.upsertChunks(spec.slug, spec.chunks, chunkOpts);
+        await tx.upsertChunks(spec.slug, spec.chunks, txOpts);
      } else {
        await tx.deleteChunks(spec.slug, txOpts);
      }
@@ -35,7 +35,6 @@ import { tryAcquireDbLock } from '../../db-lock.ts';
 import { BudgetTracker, BudgetExhausted } from '../../budget/budget-tracker.ts';
 import { withBudgetTracker } from '../../ai/gateway.ts';
 import { embedStaleForSource } from '../../embed-stale.ts';
-import { resolveWriteColumnForEngine } from '../../search/embedding-column.ts';
 import { currentEmbeddingSignature } from '../../embedding.ts';
 import { type DbPacer, createDbPacer, createNoopPacer } from '../../db-pacer.ts';
 import { resolvePaceMode, loadPaceModeConfig, readPaceEnv } from '../../pace-mode.ts';
@@ -165,16 +164,12 @@ export function makeEmbedBackfillHandler(engine: BrainEngine) {
    // the supervisor, so pacing it is the headline win.
    const { pacer, concurrency } = await resolveBackfillPacer(engine, job.data);

-    // #1262: resolve the write-side embedding column once at the job boundary.
-    const embeddingColumn = await resolveWriteColumnForEngine(engine);
-
    try {
      const result = await withBudgetTracker(tracker, async () =>
        embedStaleForSource(engine, sourceId, {
          batchSize,
          signal: job.signal,
          pacer,
-          ...(embeddingColumn && { embeddingColumn }),
          ...(concurrency !== undefined && { concurrency }),
          // v0.41.31: re-embed pages whose model signature drifted + stamp
          // provenance as chunks land.
@@ -40,7 +40,6 @@ import type {
  BrainStats, BrainHealth,
  IngestLogEntry, IngestLogInput,
  EngineConfig,
-  ResolvedColumn,
  EvalCandidate, EvalCandidateInput,
  EvalCaptureFailure, EvalCaptureFailureReason,
  SalienceOpts, SalienceResult, AnomaliesOpts, AnomalyResult,
@@ -1004,7 +1003,15 @@ export class PGLiteEngine implements BrainEngine {
    return { slug: r.slug, id: Number(r.id) };
  }

-  async putPage(slug: string, page: PageInput, opts?: { sourceId?: string }): Promise<Page> {
+  async putPage(
+    slug: string,
+    page: PageInput,
+    opts?: { sourceId?: string; signal?: AbortSignal },
+  ): Promise<Page> {
+    // #2750: PGLite is in-process WASM — no query cancellation. Pre-check so
+    // an already-fired deadline skips the write; abort is cooperative
+    // between calls (same posture as executeRaw's documented gap).
+    if (opts?.signal?.aborted) throw new DOMException('aborted', 'AbortError');
    slug = validateSlug(slug);
    const hash = page.content_hash || contentHash(page);
    const frontmatter = page.frontmatter || {};
@@ -2231,20 +2238,12 @@ export class PGLiteEngine implements BrainEngine {
  }

  // Chunks
-  async upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string; embeddingColumn?: ResolvedColumn } & BatchOpts): Promise<void> {
+  async upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string } & BatchOpts): Promise<void> {
    return this.batchRetry(opts?.auditSite ?? 'upsertChunks', opts?.signal, () => this._upsertChunksOnce(slug, chunks, opts), chunks.length);
  }

-  private async _upsertChunksOnce(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string; embeddingColumn?: ResolvedColumn }): Promise<void> {
+  private async _upsertChunksOnce(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string }): Promise<void> {
    const sourceId = opts?.sourceId ?? 'default';
-    // #1262: caller-resolved write target for TEXT embeddings. Descriptor
-    // names are identifier-validated + quoted by buildVectorCastFragment;
-    // omitted => legacy `embedding vector`. Mirrors postgres-engine.ts.
-    const targetFragment = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn)
-      : undefined;
-    const targetCol = targetFragment?.col ?? 'embedding';
-    const embeddingCast = targetFragment?.castSql.replace('$1::', '') ?? 'vector';

    // Source-scope the page-id lookup so duplicate slugs in different sources
    // do not return multiple rows or target the wrong page.
@@ -2279,7 +2278,7 @@ export class PGLiteEngine implements BrainEngine {
    // list. Image chunks pass embedding=null + embedding_image=Float32Array
    // (1024-dim Voyage). Text/code chunks pass embedding=Float32Array +
    // embedding_image=null. Default modality='text' when omitted.
-    const cols = `(page_id, chunk_index, chunk_text, chunk_source, ${targetCol}, model, token_count, embedded_at, language, symbol_name, symbol_type, start_line, end_line, parent_symbol_path, doc_comment, symbol_name_qualified, modality, embedding_image)`;
+    const cols = '(page_id, chunk_index, chunk_text, chunk_source, embedding, model, token_count, embedded_at, language, symbol_name, symbol_type, start_line, end_line, parent_symbol_path, doc_comment, symbol_name_qualified, modality, embedding_image)';
    const rowParts: string[] = [];
    const params: unknown[] = [];
    let paramIdx = 1;
@@ -2297,7 +2296,7 @@ export class PGLiteEngine implements BrainEngine {
      const modality = chunk.modality ?? 'text';

      // Inline ::vector NULL literals to avoid a per-branch placeholder.
-      const embeddingPh = embeddingStr ? `$${paramIdx++}::${embeddingCast}` : 'NULL';
+      const embeddingPh = embeddingStr ? `$${paramIdx++}::vector` : 'NULL';
      const embeddedAtPh = embeddingStr ? 'now()' : 'NULL';
      const embeddingImagePh = embeddingImageStr ? `$${paramIdx++}::vector` : 'NULL';

@@ -2336,19 +2335,19 @@ export class PGLiteEngine implements BrainEngine {
       ON CONFLICT (page_id, chunk_index) DO UPDATE SET
         chunk_text = EXCLUDED.chunk_text,
         chunk_source = EXCLUDED.chunk_source,
-         ${targetCol} = CASE
-           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text THEN EXCLUDED.${targetCol}
-           WHEN content_chunks.${targetCol} IS NULL THEN EXCLUDED.${targetCol}
+         embedding = CASE
+           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text THEN EXCLUDED.embedding
+           WHEN content_chunks.embedding IS NULL THEN EXCLUDED.embedding
           WHEN EXCLUDED.embedded_at IS NOT NULL
                AND (content_chunks.embedded_at IS NULL OR EXCLUDED.embedded_at > content_chunks.embedded_at)
-                THEN EXCLUDED.${targetCol}
-           ELSE content_chunks.${targetCol}
+                THEN EXCLUDED.embedding
+           ELSE content_chunks.embedding
         END,
         model = COALESCE(EXCLUDED.model, content_chunks.model),
         token_count = EXCLUDED.token_count,
         embedded_at = CASE
-           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text AND EXCLUDED.${targetCol} IS NULL THEN NULL
-           WHEN content_chunks.${targetCol} IS NULL AND EXCLUDED.${targetCol} IS NOT NULL THEN EXCLUDED.embedded_at
+           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text AND EXCLUDED.embedding IS NULL THEN NULL
+           WHEN content_chunks.embedding IS NULL AND EXCLUDED.embedding IS NOT NULL THEN EXCLUDED.embedded_at
           WHEN EXCLUDED.embedded_at IS NOT NULL
                AND (content_chunks.embedded_at IS NULL OR EXCLUDED.embedded_at > content_chunks.embedded_at)
                THEN EXCLUDED.embedded_at
@@ -2386,19 +2385,14 @@ export class PGLiteEngine implements BrainEngine {
   * drift (NULL grandfathered → never stale). Shared by countStaleChunks +
   * sumStaleChunkChars so they can't drift.
   */
-  private buildStaleChunkWhere(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): { where: string; params: unknown[] } {
-    // #1262: staleness targets the caller-resolved write column when set
-    // (identifier-validated + quoted); legacy `embedding` otherwise.
-    const staleCol = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn).col
-      : 'embedding';
+  private buildStaleChunkWhere(opts?: { sourceId?: string; signature?: string }): { where: string; params: unknown[] } {
    const params: unknown[] = [];
    const conds: string[] = [];
    if (opts?.signature !== undefined) {
      params.push(opts.signature);
-      conds.push(`(cc.${staleCol} IS NULL OR (p.embedding_signature IS NOT NULL AND p.embedding_signature <> $${params.length}))`);
+      conds.push(`(cc.embedding IS NULL OR (p.embedding_signature IS NOT NULL AND p.embedding_signature <> $${params.length}))`);
    } else {
-      conds.push(`cc.${staleCol} IS NULL`);
+      conds.push(`cc.embedding IS NULL`);
    }
    conds.push(`NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')`);
    if (opts?.sourceId !== undefined) {
@@ -2408,7 +2402,7 @@ export class PGLiteEngine implements BrainEngine {
    return { where: conds.join(' AND '), params };
  }

-  async countStaleChunks(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number> {
+  async countStaleChunks(opts?: { sourceId?: string; signature?: string }): Promise<number> {
    // D7: source-scoped count for `gbrain embed --stale --source X`. Always
    // JOIN pages so embed-skip + signature predicates apply. PGLite is
    // PostgreSQL 17.5 in WASM and supports the full JSONB operator set.
@@ -2424,7 +2418,7 @@ export class PGLiteEngine implements BrainEngine {
    return Number(count);
  }

-  async sumStaleChunkChars(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number> {
+  async sumStaleChunkChars(opts?: { sourceId?: string; signature?: string }): Promise<number> {
    // Sibling of countStaleChunks: same stale predicate, summing chunk_text
    // length for the sync cost preview. ::bigint guards int4 overflow.
    const { where, params } = this.buildStaleChunkWhere(opts);
@@ -2477,17 +2471,11 @@ export class PGLiteEngine implements BrainEngine {
    sourceId?: string;
    orderBy?: 'page_id' | 'updated_desc';
    afterUpdatedAt?: string | null;
-    embeddingColumn?: ResolvedColumn;
  }): Promise<StaleChunkRow[]> {
    const limit = opts?.batchSize ?? 2000;
    const afterPid = opts?.afterPageId ?? 0;
    const afterIdx = opts?.afterChunkIndex ?? -1;
    const orderBy = opts?.orderBy ?? 'page_id';
-    // #1262: staleness follows the caller-resolved write column (validated +
-    // quoted identifier); legacy `embedding` otherwise.
-    const staleCol = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn).col
-      : 'embedding';

    // v0.41.18.0 (A13, codex #9): --priority recent path. See postgres-engine
    // sibling for full rationale. Same composite cursor + ORDER BY.
@@ -2501,7 +2489,7 @@ export class PGLiteEngine implements BrainEngine {
                  p.updated_at
             FROM content_chunks cc
             JOIN pages p ON p.id = cc.page_id
-            WHERE cc.${staleCol} IS NULL
+            WHERE cc.embedding IS NULL
              AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            ORDER BY p.updated_at DESC NULLS LAST, p.id ASC, cc.chunk_index ASC
            LIMIT $1`,
@@ -2512,7 +2500,7 @@ export class PGLiteEngine implements BrainEngine {
                  p.updated_at
             FROM content_chunks cc
             JOIN pages p ON p.id = cc.page_id
-            WHERE cc.${staleCol} IS NULL
+            WHERE cc.embedding IS NULL
              AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
              AND (
                p.updated_at < $1::timestamptz
@@ -2531,7 +2519,7 @@ export class PGLiteEngine implements BrainEngine {
                p.updated_at
           FROM content_chunks cc
           JOIN pages p ON p.id = cc.page_id
-          WHERE cc.${staleCol} IS NULL
+          WHERE cc.embedding IS NULL
            AND p.source_id = $1
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
          ORDER BY p.updated_at DESC NULLS LAST, p.id ASC, cc.chunk_index ASC
@@ -2543,7 +2531,7 @@ export class PGLiteEngine implements BrainEngine {
                p.updated_at
           FROM content_chunks cc
           JOIN pages p ON p.id = cc.page_id
-          WHERE cc.${staleCol} IS NULL
+          WHERE cc.embedding IS NULL
            AND p.source_id = $1
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            AND (
@@ -2568,7 +2556,7 @@ export class PGLiteEngine implements BrainEngine {
                cc.model, cc.token_count, p.source_id, cc.page_id
           FROM content_chunks cc
           JOIN pages p ON p.id = cc.page_id
-          WHERE cc.${staleCol} IS NULL
+          WHERE cc.embedding IS NULL
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            AND (cc.page_id, cc.chunk_index) > ($1, $2)
          ORDER BY cc.page_id, cc.chunk_index
@@ -2582,7 +2570,7 @@ export class PGLiteEngine implements BrainEngine {
              cc.model, cc.token_count, p.source_id, cc.page_id
         FROM content_chunks cc
         JOIN pages p ON p.id = cc.page_id
-        WHERE cc.${staleCol} IS NULL
+        WHERE cc.embedding IS NULL
          AND p.source_id = $1
          AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
          AND (cc.page_id, cc.chunk_index) > ($2, $3)
@@ -50,7 +50,6 @@ import type {
  BrainStats, BrainHealth,
  IngestLogEntry, IngestLogInput,
  EngineConfig,
-  ResolvedColumn,
  EvalCandidate, EvalCandidateInput,
  EvalCaptureFailure, EvalCaptureFailureReason,
  SalienceOpts, SalienceResult, AnomaliesOpts, AnomalyResult,
@@ -73,6 +72,32 @@ function escapeSqlStringLiteral(value: string): string {
  return value.replace(/'/g, "''");
 }

+/**
+ * #2750: race a promise against an AbortSignal, detaching the listener once
+ * settled (long-lived drain signals are reused across many calls, so a bare
+ * Promise.race would leak one listener per call). The abandoned promise keeps
+ * running; used only for pool-acquisition waits where that is harmless.
+ */
+function waitForSignal<T>(work: Promise<T>, signal?: AbortSignal): Promise<T> {
+  if (!signal) return work;
+  if (signal.aborted) return Promise.reject(new DOMException('aborted', 'AbortError'));
+  return new Promise<T>((resolve, reject) => {
+    let settled = false;
+    const finish = (fn: () => void) => {
+      if (settled) return;
+      settled = true;
+      signal.removeEventListener('abort', onAbort);
+      fn();
+    };
+    const onAbort = () => finish(() => reject(new DOMException('aborted', 'AbortError')));
+    signal.addEventListener('abort', onAbort, { once: true });
+    work.then(
+      (value) => finish(() => resolve(value)),
+      (err) => finish(() => reject(err)),
+    );
+  });
+}
+
 export function getPostgresSchema(
  dims: number = DEFAULT_EMBEDDING_DIMENSIONS,
  model: string = DEFAULT_EMBEDDING_MODEL,
@@ -1062,7 +1087,12 @@ export class PostgresEngine implements BrainEngine {
    });
  }

-  async putPage(slug: string, page: PageInput, opts?: { sourceId?: string }): Promise<Page> {
+  async putPage(
+    slug: string,
+    page: PageInput,
+    opts?: { sourceId?: string; signal?: AbortSignal },
+  ): Promise<Page> {
+    if (opts?.signal?.aborted) throw new DOMException('aborted', 'AbortError');
    slug = validateSlug(slug);
    const sql = this.sql;
    const hash = page.content_hash || contentHash(page);
@@ -1097,7 +1127,7 @@ export class PostgresEngine implements BrainEngine {
    const sourceUri = page.source_uri ?? null;
    const ingestedVia = page.ingested_via ?? null;
    const ingestedAt = (sourceKind || sourceUri || ingestedVia) ? new Date() : null;
-    const rows = await sql`
+    const pending = sql`
      INSERT INTO pages (source_id, slug, type, page_kind, title, compiled_truth, timeline, frontmatter, content_hash, updated_at, effective_date, effective_date_source, import_filename, chunker_version, source_path, source_kind, source_uri, ingested_via, ingested_at)
      VALUES (${sourceId}, ${slug}, ${page.type}, ${pageKind}, ${page.title}, ${page.compiled_truth}, ${page.timeline || ''}, ${sql.json(frontmatter as Parameters<typeof sql.json>[0])}, ${hash}, now(), ${effectiveDate}, ${effectiveDateSource}, ${importFilename}, COALESCE(${chunkerVersion}::smallint, ${MARKDOWN_CHUNKER_VERSION}), ${sourcePath}, ${sourceKind}, ${sourceUri}, ${ingestedVia}, ${ingestedAt})
      ON CONFLICT (source_id, slug) DO UPDATE SET
@@ -1120,6 +1150,22 @@ export class PostgresEngine implements BrainEngine {
        ingested_at           = COALESCE(EXCLUDED.ingested_at,           pages.ingested_at)
      RETURNING id, source_id, slug, type, title, compiled_truth, timeline, frontmatter, content_hash, created_at, updated_at, effective_date, effective_date_source, import_filename, source_kind, source_uri, ingested_via, ingested_at
    `;
+    // #2750: cancel the in-flight statement when the caller's deadline fires,
+    // same .cancel() wiring as runUnsafe (postgres.js pending queries).
+    if (opts?.signal) {
+      const signal = opts.signal;
+      const onAbort = () => {
+        try { (pending as unknown as { cancel?: () => void }).cancel?.(); } catch { /* best-effort */ }
+      };
+      signal.addEventListener('abort', onAbort, { once: true });
+      try {
+        const rows = await pending;
+        return rowToPage(rows[0]);
+      } finally {
+        signal.removeEventListener('abort', onAbort);
+      }
+    }
+    const rows = await pending;
    return rowToPage(rows[0]);
  }

@@ -2381,21 +2427,13 @@ export class PostgresEngine implements BrainEngine {
  }

  // Chunks
-  async upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string; embeddingColumn?: ResolvedColumn } & BatchOpts): Promise<void> {
+  async upsertChunks(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string } & BatchOpts): Promise<void> {
    return this.batchRetry(opts?.auditSite ?? 'upsertChunks', opts?.signal, () => this._upsertChunksOnce(slug, chunks, opts), chunks.length);
  }

-  private async _upsertChunksOnce(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string; embeddingColumn?: ResolvedColumn }): Promise<void> {
+  private async _upsertChunksOnce(slug: string, chunks: ChunkInput[], opts?: { sourceId?: string }): Promise<void> {
    const sql = this.sql;
    const sourceId = opts?.sourceId ?? 'default';
-    // #1262: caller-resolved write target for TEXT embeddings. Descriptor
-    // names are identifier-validated + quoted by buildVectorCastFragment;
-    // omitted => legacy `embedding vector`.
-    const targetFragment = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn)
-      : undefined;
-    const targetCol = targetFragment?.col ?? 'embedding';
-    const embeddingCast = targetFragment?.castSql.replace('$1::', '') ?? 'vector';

    // Source-scope the page-id lookup. Without this filter, multi-source
    // brains where the slug exists in 2+ sources return >1 row and the
@@ -2422,7 +2460,7 @@ export class PostgresEngine implements BrainEngine {
    // scope metadata through upserts.
    // v0.27.1 (Phase 8): added `modality` + `embedding_image` to the column
    // list. Image chunks pass embedding=null + embedding_image=Float32Array.
-    const cols = `(page_id, chunk_index, chunk_text, chunk_source, ${targetCol}, model, token_count, embedded_at, language, symbol_name, symbol_type, start_line, end_line, parent_symbol_path, doc_comment, symbol_name_qualified, modality, embedding_image)`;
+    const cols = '(page_id, chunk_index, chunk_text, chunk_source, embedding, model, token_count, embedded_at, language, symbol_name, symbol_type, start_line, end_line, parent_symbol_path, doc_comment, symbol_name_qualified, modality, embedding_image)';
    const rows: string[] = [];
    const params: unknown[] = [];
    let paramIdx = 1;
@@ -2439,7 +2477,7 @@ export class PostgresEngine implements BrainEngine {
        : null;
      const modality = chunk.modality ?? 'text';

-      const embeddingPh = embeddingStr ? `$${paramIdx++}::${embeddingCast}` : 'NULL';
+      const embeddingPh = embeddingStr ? `$${paramIdx++}::vector` : 'NULL';
      const embeddedAtPh = embeddingStr ? 'now()' : 'NULL';
      const embeddingImagePh = embeddingImageStr ? `$${paramIdx++}::vector` : 'NULL';

@@ -2487,19 +2525,19 @@ export class PostgresEngine implements BrainEngine {
       ON CONFLICT (page_id, chunk_index) DO UPDATE SET
         chunk_text = EXCLUDED.chunk_text,
         chunk_source = EXCLUDED.chunk_source,
-         ${targetCol} = CASE
-           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text THEN EXCLUDED.${targetCol}
-           WHEN content_chunks.${targetCol} IS NULL THEN EXCLUDED.${targetCol}
+         embedding = CASE
+           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text THEN EXCLUDED.embedding
+           WHEN content_chunks.embedding IS NULL THEN EXCLUDED.embedding
           WHEN EXCLUDED.embedded_at IS NOT NULL
                AND (content_chunks.embedded_at IS NULL OR EXCLUDED.embedded_at > content_chunks.embedded_at)
-                THEN EXCLUDED.${targetCol}
-           ELSE content_chunks.${targetCol}
+                THEN EXCLUDED.embedding
+           ELSE content_chunks.embedding
         END,
         model = COALESCE(EXCLUDED.model, content_chunks.model),
         token_count = EXCLUDED.token_count,
         embedded_at = CASE
-           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text AND EXCLUDED.${targetCol} IS NULL THEN NULL
-           WHEN content_chunks.${targetCol} IS NULL AND EXCLUDED.${targetCol} IS NOT NULL THEN EXCLUDED.embedded_at
+           WHEN EXCLUDED.chunk_text != content_chunks.chunk_text AND EXCLUDED.embedding IS NULL THEN NULL
+           WHEN content_chunks.embedding IS NULL AND EXCLUDED.embedding IS NOT NULL THEN EXCLUDED.embedded_at
           WHEN EXCLUDED.embedded_at IS NOT NULL
                AND (content_chunks.embedded_at IS NULL OR EXCLUDED.embedded_at > content_chunks.embedded_at)
                THEN EXCLUDED.embedded_at
@@ -2539,19 +2577,14 @@ export class PostgresEngine implements BrainEngine {
   * embedding_signature drift (NULL grandfathered). Shared by
   * countStaleChunks + sumStaleChunkChars (parity with the PGLite sibling).
   */
-  private buildStaleChunkWhere(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): { where: string; params: unknown[] } {
-    // #1262: staleness targets the caller-resolved write column when set
-    // (identifier-validated + quoted); legacy `embedding` otherwise.
-    const staleCol = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn).col
-      : 'embedding';
+  private buildStaleChunkWhere(opts?: { sourceId?: string; signature?: string }): { where: string; params: unknown[] } {
    const params: unknown[] = [];
    const conds: string[] = [];
    if (opts?.signature !== undefined) {
      params.push(opts.signature);
-      conds.push(`(cc.${staleCol} IS NULL OR (p.embedding_signature IS NOT NULL AND p.embedding_signature <> $${params.length}))`);
+      conds.push(`(cc.embedding IS NULL OR (p.embedding_signature IS NOT NULL AND p.embedding_signature <> $${params.length}))`);
    } else {
-      conds.push(`cc.${staleCol} IS NULL`);
+      conds.push(`cc.embedding IS NULL`);
    }
    conds.push(`NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')`);
    if (opts?.sourceId !== undefined) {
@@ -2561,7 +2594,7 @@ export class PostgresEngine implements BrainEngine {
    return { where: conds.join(' AND '), params };
  }

-  async countStaleChunks(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number> {
+  async countStaleChunks(opts?: { sourceId?: string; signature?: string }): Promise<number> {
    // Always JOIN pages so the embed_skip + signature predicates apply.
    // D7: source_id scoping. v0.41.31: optional signature widens staleness
    // to embedding_signature drift (NULL grandfathered).
@@ -2579,7 +2612,7 @@ export class PostgresEngine implements BrainEngine {
    });
  }

-  async sumStaleChunkChars(opts?: { sourceId?: string; signature?: string; embeddingColumn?: ResolvedColumn }): Promise<number> {
+  async sumStaleChunkChars(opts?: { sourceId?: string; signature?: string }): Promise<number> {
    // Sibling of countStaleChunks: same stale predicate, summing chunk_text
    // length for the sync cost preview. ::bigint guards int4 overflow.
    const { where, params } = this.buildStaleChunkWhere(opts);
@@ -2632,18 +2665,11 @@ export class PostgresEngine implements BrainEngine {
    sourceId?: string;
    orderBy?: 'page_id' | 'updated_desc';
    afterUpdatedAt?: string | null;
-    embeddingColumn?: ResolvedColumn;
  }): Promise<StaleChunkRow[]> {
    const limit = opts?.batchSize ?? 2000;
    const afterPid = opts?.afterPageId ?? 0;
    const afterIdx = opts?.afterChunkIndex ?? -1;
    const orderBy = opts?.orderBy ?? 'page_id';
-    // #1262: staleness follows the caller-resolved write column (validated +
-    // quoted identifier); legacy `embedding` otherwise. Interpolated below as
-    // an unsafe FRAGMENT (identifiers can't be bound parameters).
-    const staleCol = opts?.embeddingColumn
-      ? buildVectorCastFragment(opts.embeddingColumn).col
-      : 'embedding';

    // RLS scope binding (opt-in via GBRAIN_RLS_SCOPE_BINDING).
    return await this.withScopedReadTransaction(undefined, opts?.sourceId, async (tx) => {
@@ -2660,7 +2686,7 @@ export class PostgresEngine implements BrainEngine {
                   p.updated_at
            FROM content_chunks cc
            JOIN pages p ON p.id = cc.page_id
-            WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+            WHERE cc.embedding IS NULL
              AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            ORDER BY p.updated_at DESC NULLS LAST, p.id ASC, cc.chunk_index ASC
            LIMIT ${limit}
@@ -2670,7 +2696,7 @@ export class PostgresEngine implements BrainEngine {
                   p.updated_at
            FROM content_chunks cc
            JOIN pages p ON p.id = cc.page_id
-            WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+            WHERE cc.embedding IS NULL
              AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
              AND (
                p.updated_at < ${afterUpdated}::timestamptz
@@ -2688,7 +2714,7 @@ export class PostgresEngine implements BrainEngine {
                 p.updated_at
          FROM content_chunks cc
          JOIN pages p ON p.id = cc.page_id
-          WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+          WHERE cc.embedding IS NULL
            AND p.source_id = ${opts.sourceId}
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
          ORDER BY p.updated_at DESC NULLS LAST, p.id ASC, cc.chunk_index ASC
@@ -2699,7 +2725,7 @@ export class PostgresEngine implements BrainEngine {
                 p.updated_at
          FROM content_chunks cc
          JOIN pages p ON p.id = cc.page_id
-          WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+          WHERE cc.embedding IS NULL
            AND p.source_id = ${opts.sourceId}
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            AND (
@@ -2719,7 +2745,7 @@ export class PostgresEngine implements BrainEngine {
                 cc.model, cc.token_count, p.source_id, cc.page_id
          FROM content_chunks cc
          JOIN pages p ON p.id = cc.page_id
-          WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+          WHERE cc.embedding IS NULL
            AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
            AND (cc.page_id, cc.chunk_index) > (${afterPid}, ${afterIdx})
          ORDER BY cc.page_id, cc.chunk_index
@@ -2732,7 +2758,7 @@ export class PostgresEngine implements BrainEngine {
               cc.model, cc.token_count, p.source_id, cc.page_id
        FROM content_chunks cc
        JOIN pages p ON p.id = cc.page_id
-        WHERE ${tx.unsafe(`cc.${staleCol} IS NULL`)}
+        WHERE cc.embedding IS NULL
          AND p.source_id = ${opts.sourceId}
          AND NOT (COALESCE(p.frontmatter, '{}'::jsonb) ? 'embed_skip')
          AND (cc.page_id, cc.chunk_index) > (${afterPid}, ${afterIdx})
@@ -5828,11 +5854,16 @@ export class PostgresEngine implements BrainEngine {
    params?: unknown[],
    opts?: { signal?: AbortSignal },
  ): Promise<T[]> {
+    // #2750: an already-fired signal short-circuits BEFORE any pool routing,
+    // and the direct-pool acquisition itself is signal-bounded — under pooler
+    // exhaustion `ddl()` can stall indefinitely, which used to make even a
+    // "bounded" lock release hang past its caller's deadline.
+    if (opts?.signal?.aborted) throw new DOMException('aborted', 'AbortError');
    // Inside an open transaction, _sql is the reserved tx connection (set via
    // defineProperty in transaction()); never reroute off it.
    const inTransaction = this._sql !== null && this.connectionManager?.peekReadPool() !== this._sql;
    const conn = (!inTransaction && this.connectionManager?.isDualPoolActive())
-      ? await this.connectionManager.ddl()
+      ? await waitForSignal(this.connectionManager.ddl(), opts?.signal)
      : this.sql;
    return this.runUnsafe<T>(conn, sql, params, opts);
  }
@@ -443,80 +443,6 @@ export function resolveEmbeddingColumn(
  };
 }

-/**
- * Resolves the WRITE-side embedding column for the currently configured
- * embedding model (#1262). The read-side resolver above answers "which
- * column does this query search?"; this one answers "which column should
- * newly produced text embeddings land in?".
- *
- * Unlike read-side search, writes take no per-call column override. The
- * import/embed boundary resolves once from merged config + gateway state
- * and passes the descriptor into `engine.upsertChunks`; engines stay
- * config-free (same contract as the read-side descriptor).
- *
- * Behavior:
- *   - no user-declared `embedding_columns` => undefined (legacy brain,
- *     writes keep targeting the default `embedding` column)
- *   - a user-declared entry whose `provider` matches the current
- *     embedding model => that entry's descriptor
- *   - no provider match => undefined (fall back to legacy `embedding`)
- *
- * Only USER-declared entries are consulted — never the cfg-derived
- * builtins. The `embedding_image` builtin's provider is the multimodal
- * model; matching it here would misroute text embeddings into the image
- * column. The no-match fallback is intentional: switching models before
- * registering a matching column must not silently write vectors into an
- * arbitrary column.
- */
-export function resolveWriteColumn(cfg: GBrainConfig): ResolvedColumn | undefined {
-  const userColumns = cfg.embedding_columns;
-  if (
-    !userColumns ||
-    typeof userColumns !== 'object' ||
-    Array.isArray(userColumns) ||
-    Object.keys(userColumns).length === 0
-  ) {
-    return undefined;
-  }
-
-  // Same model-resolution chain as the registry builtin: cfg > gateway > default.
-  let gwModel: string | undefined;
-  try {
-    const gw = require('../ai/gateway.ts') as typeof import('../ai/gateway.ts');
-    gwModel = gw.getEmbeddingModel();
-  } catch {
-    // Gateway unconfigured — fall through to the canonical default.
-  }
-  const currentModel = cfg.embedding_model ?? gwModel ?? DEFAULT_EMBEDDING_MODEL;
-
-  for (const [name, entry] of Object.entries(userColumns)) {
-    if (!entry) continue;
-    validateColumnKey(name);
-    validateColumnConfig(name, entry);
-    if (entry.provider !== currentModel) continue;
-    return {
-      name,
-      type: entry.type,
-      dimensions: entry.dimensions,
-      embeddingModel: entry.provider,
-    };
-  }
-  return undefined;
-}
-
-/**
- * Engine-boundary convenience: merged config (file/env + DB plane) →
- * resolveWriteColumn. Dynamic import keeps config.ts out of this module's
- * static graph (mirrors the gateway require above).
- */
-export async function resolveWriteColumnForEngine(
-  engine: { getConfig(key: string): Promise<string | null | undefined> },
-): Promise<ResolvedColumn | undefined> {
-  const { loadConfigWithEngine } = await import('../config.ts');
-  const cfg = await loadConfigWithEngine(engine);
-  return cfg ? resolveWriteColumn(cfg) : undefined;
-}
-
 /**
 * True when the resolved column is the default `embedding` name.
 * Name-based check; does not compare embedding space.
@@ -17,6 +17,7 @@ import { runPhaseExtractAtoms, parseAtomsResponse } from '../../src/core/cycle/e
 import { runPhaseSynthesizeConcepts } from '../../src/core/cycle/synthesize-concepts.ts';
 import { resetPgliteState } from '../helpers/reset-pglite.ts';
 import type { ChatResult, ChatOpts } from '../../src/core/ai/gateway.ts';
+import type { BrainEngine } from '../../src/core/engine.ts';

 let engine: PGLiteEngine;

@@ -177,6 +178,180 @@ describe('v0.41 T5: runPhaseExtractAtoms via stubbed chat', () => {
    expect((result.details?.failures as unknown[]).length).toBe(1);
  });

+  // ── #2750: caller deadline bounds the phase ────────────────────────────
+
+  test('caller deadline aborts a hung chat before processing the next item', async () => {
+    let calls = 0;
+    const chat = async (opts: ChatOpts) => {
+      calls++;
+      return await new Promise<never>((_resolve, reject) => {
+        const signal = opts.abortSignal;
+        if (!signal) return reject(new Error('missing abort signal'));
+        if (signal.aborted) return reject(signal.reason);
+        signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+      });
+    };
+    const started = Date.now();
+    const result = await runPhaseExtractAtoms(engine, {
+      _transcripts: [
+        { filePath: '/hung.txt', content: 'a', contentHash: 'hung-a' },
+        { filePath: '/never.txt', content: 'b', contentHash: 'hung-b' },
+      ],
+      _pages: [],
+      _chat: chat as typeof import('../../src/core/ai/gateway.ts').chat,
+      abortSignal: AbortSignal.timeout(25),
+    });
+    expect(Date.now() - started).toBeLessThan(2_000);
+    expect(calls).toBe(1);
+    expect(result.status).toBe('ok');
+    expect(result.details?.deadline_aborted).toBe(true);
+    expect(result.details?.atoms_extracted).toBe(0);
+    expect(result.details?.failures).toEqual([]);
+  });
+
+  test('billable chat usage is counted when the deadline fires as the response resolves', async () => {
+    const controller = new AbortController();
+    const chat = async (opts: ChatOpts): Promise<ChatResult> => {
+      controller.abort(new DOMException('deadline', 'TimeoutError'));
+      return stubChat(`[{"title":"late","atom_type":"insight","body":"b"}]`, {
+        input_tokens: 1_000,
+        output_tokens: 500,
+      })(opts);
+    };
+    const result = await runPhaseExtractAtoms(engine, {
+      _transcripts: [{ filePath: '/late.txt', content: 'a', contentHash: 'late' }],
+      _pages: [],
+      _chat: chat,
+      abortSignal: controller.signal,
+    });
+    expect(result.details?.deadline_aborted).toBe(true);
+    expect(Number(result.details?.estimated_spend_usd)).toBeGreaterThan(0);
+    expect(result.details?.atoms_extracted).toBe(0);
+  });
+
+  test('deadline after partial progress still writes receipt and incomplete rollup', async () => {
+    const controller = new AbortController();
+    let calls = 0;
+    let notifySecondChat!: () => void;
+    const secondChatStarted = new Promise<void>((resolve) => { notifySecondChat = resolve; });
+    const chat = async (opts: ChatOpts): Promise<ChatResult> => {
+      calls++;
+      if (calls === 1) {
+        return stubChat(`[{"title":"committed","atom_type":"insight","body":"b"}]`)(opts);
+      }
+      notifySecondChat();
+      return await new Promise<never>((_resolve, reject) => {
+        const signal = opts.abortSignal;
+        if (!signal) return reject(new Error('missing abort signal'));
+        signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+      });
+    };
+
+    const pending = runPhaseExtractAtoms(engine, {
+      _transcripts: [
+        { filePath: '/committed.txt', content: 'a', contentHash: 'committed-a' },
+        { filePath: '/hung.txt', content: 'b', contentHash: 'hung-b' },
+      ],
+      _pages: [],
+      _chat: chat,
+      abortSignal: controller.signal,
+    });
+    await secondChatStarted;
+    controller.abort(new DOMException('deadline', 'TimeoutError'));
+    const result = await pending;
+
+    const atoms = await engine.executeRaw<{ n: number }>(
+      `SELECT COUNT(*)::int AS n FROM pages WHERE type = 'atom'`,
+    );
+    const receipts = await engine.executeRaw<{ n: number }>(
+      `SELECT COUNT(*)::int AS n FROM pages WHERE type = 'extract_receipt'`,
+    );
+    const rollups = await engine.executeRaw<{
+      cost_usd: string | number;
+      round_completed_count: string | number;
+    }>(
+      `SELECT cost_usd, round_completed_count
+         FROM extract_rollup_7d
+        WHERE kind = 'atoms' AND source_id = 'default'`,
+    );
+    expect(result.details?.deadline_aborted).toBe(true);
+    expect(atoms[0].n).toBe(1);
+    expect(receipts[0].n).toBe(1);
+    expect(Number(rollups[0].cost_usd)).toBeGreaterThan(0);
+    expect(Number(rollups[0].round_completed_count)).toBe(0);
+  });
+
+  test('bookkeeping runs on a fresh grace signal, not the fired work deadline', async () => {
+    const controller = new AbortController();
+    let putCalls = 0;
+    let receiptSignal: AbortSignal | undefined;
+    let rollupSignal: AbortSignal | undefined;
+    const signalAwareEngine = {
+      executeRaw: async (sql: string, _params?: unknown[], opts?: { signal?: AbortSignal }) => {
+        if (sql.includes('INSERT INTO extract_rollup_7d')) rollupSignal = opts?.signal;
+        return [];
+      },
+      putPage: async (_slug: string, _page: unknown, opts?: { signal?: AbortSignal }) => {
+        putCalls++;
+        if (putCalls === 1) {
+          // Atom write in flight; the work deadline fires before bookkeeping.
+          controller.abort(new DOMException('work deadline', 'TimeoutError'));
+        } else {
+          receiptSignal = opts?.signal;
+        }
+        return {};
+      },
+    } as unknown as BrainEngine;
+
+    const result = await runPhaseExtractAtoms(signalAwareEngine, {
+      _transcripts: [{ filePath: '/one.txt', content: 'a', contentHash: 'one' }],
+      _pages: [],
+      _chat: stubChat(`[{"title":"one","atom_type":"insight","body":"b"}]`),
+      abortSignal: controller.signal,
+    });
+
+    expect(result.details?.atoms_extracted).toBe(1);
+    expect(putCalls).toBe(2); // atom write + receipt write
+    expect(receiptSignal).toBeDefined();
+    expect(receiptSignal).not.toBe(controller.signal);
+    expect(receiptSignal?.aborted).toBe(false);
+    expect(rollupSignal).toBe(receiptSignal);
+  });
+
+  test('caller deadline cancels a hung atom write and stops the phase', async () => {
+    const controller = new AbortController();
+    let notifyWriteStarted!: () => void;
+    const writeStarted = new Promise<void>((resolve) => { notifyWriteStarted = resolve; });
+    let writeCalls = 0;
+    const signalAwareEngine = {
+      executeRaw: async () => [],
+      putPage: async (_slug: string, _page: unknown, opts?: { signal?: AbortSignal }) => {
+        writeCalls++;
+        notifyWriteStarted();
+        return await new Promise<never>((_resolve, reject) => {
+          const signal = opts?.signal;
+          if (!signal) return reject(new Error('missing abort signal'));
+          if (signal.aborted) return reject(signal.reason);
+          signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+        });
+      },
+    } as unknown as BrainEngine;
+
+    const pending = runPhaseExtractAtoms(signalAwareEngine, {
+      _transcripts: [{ filePath: '/hung-write.txt', content: 'a', contentHash: 'hung-write' }],
+      _pages: [],
+      _chat: stubChat(`[{"title":"hung write","atom_type":"insight","body":"b"}]`),
+      abortSignal: controller.signal,
+    });
+    await writeStarted;
+    controller.abort(new DOMException('deadline', 'TimeoutError'));
+    const result = await pending;
+
+    expect(writeCalls).toBe(1);
+    expect(result.details?.deadline_aborted).toBe(true);
+    expect(result.details?.atoms_extracted).toBe(0);
+  });
+
  // v0.41.2.1 regression case (D9 #14 wording): with _pages:[] and same
  // _transcripts, all PRE-EXISTING PhaseResult.details fields match
  // pre-fix values byte-for-byte. The new fields (pages_processed,
@@ -241,136 +241,3 @@ describe('buildVectorCastFragment — engine SQL composer (D3)', () => {
    expect(castSql).toBe('$1::halfvec(2560)');
  });
 });
-
-describe('PGLite engine: upsertChunks write-side ResolvedColumn descriptor (#1262)', () => {
-  test('halfvec descriptor writes the text embedding to the alternate column, not legacy embedding', async () => {
-    await engine.putPage('docs/write-alt-pglite', {
-      type: 'concept',
-      title: 'Write alt column PGLite',
-      compiled_truth: 'PGLite write-side alternate embedding column test.',
-    });
-
-    const descriptor: ResolvedColumn = {
-      name: 'embedding_ze',
-      type: 'halfvec',
-      dimensions: 2560,
-      embeddingModel: 'zeroentropyai:zembed-1',
-    };
-    await engine.upsertChunks('docs/write-alt-pglite', [
-      {
-        chunk_index: 0,
-        chunk_text: 'PGLite write-side alternate embedding column test.',
-        chunk_source: 'compiled_truth',
-        embedding: new Float32Array(2560).fill(0.25),
-      },
-    ], { embeddingColumn: descriptor });
-
-    const rows = await engine.executeRaw<{
-      has_default: boolean;
-      has_ze: boolean;
-      has_embedded_at: boolean;
-    }>(
-      `SELECT embedding IS NOT NULL AS has_default,
-              embedding_ze IS NOT NULL AS has_ze,
-              embedded_at IS NOT NULL AS has_embedded_at
-         FROM content_chunks cc
-         JOIN pages p ON p.id = cc.page_id
-        WHERE p.slug = 'docs/write-alt-pglite'`,
-    );
-    expect(rows.length).toBe(1);
-    expect(rows[0].has_default).toBe(false);
-    expect(rows[0].has_ze).toBe(true);
-    expect(rows[0].has_embedded_at).toBe(true);
-  });
-
-  test('text-unchanged re-upsert without a vector preserves the alternate-column embedding', async () => {
-    const descriptor: ResolvedColumn = {
-      name: 'embedding_ze',
-      type: 'halfvec',
-      dimensions: 2560,
-      embeddingModel: 'zeroentropyai:zembed-1',
-    };
-    // Same chunk_text, no embedding: the ON CONFLICT CASE must keep the
-    // existing alternate-column vector (D24 semantics follow the column).
-    await engine.upsertChunks('docs/write-alt-pglite', [
-      {
-        chunk_index: 0,
-        chunk_text: 'PGLite write-side alternate embedding column test.',
-        chunk_source: 'compiled_truth',
-      },
-    ], { embeddingColumn: descriptor });
-    const rows = await engine.executeRaw<{ has_ze: boolean }>(
-      `SELECT embedding_ze IS NOT NULL AS has_ze
-         FROM content_chunks cc
-         JOIN pages p ON p.id = cc.page_id
-        WHERE p.slug = 'docs/write-alt-pglite'`,
-    );
-    expect(rows).toEqual([{ has_ze: true }]);
-  });
-});
-
-describe('PGLite: embed --stale converges on an alt-column brain (#1262)', () => {
-  test('boundary resolves the write column; stale scan does not re-select embedded rows', async () => {
-    const { runEmbedCore } = await import('../../src/commands/embed.ts');
-    const local = new PGLiteEngine();
-    const previousHome = process.env.GBRAIN_HOME;
-    process.env.GBRAIN_HOME = `/tmp/gbrain-write-col-stale-${Date.now()}`;
-    try {
-      await local.connect({});
-      await local.initSchema();
-      await (local as any).db.exec(
-        `ALTER TABLE content_chunks ADD COLUMN IF NOT EXISTS embedding_ze halfvec(2560)`,
-      );
-
-      const descriptor: ResolvedColumn = {
-        name: 'embedding_ze',
-        type: 'halfvec',
-        dimensions: 2560,
-        embeddingModel: 'zeroentropyai:zembed-1',
-      };
-      await local.setConfig('embedding_columns', JSON.stringify({
-        embedding_ze: { provider: 'zeroentropyai:zembed-1', dimensions: 2560, type: 'halfvec' },
-      }));
-      configureGateway({
-        embedding_model: 'zeroentropyai:zembed-1',
-        embedding_dimensions: 2560,
-        env: {},
-      });
-
-      await local.putPage('docs/stale-alt-pglite', {
-        type: 'concept',
-        title: 'Dynamic stale column',
-        compiled_truth: 'A chunk that is embedded only in the dynamic column.',
-      });
-      await local.upsertChunks('docs/stale-alt-pglite', [
-        {
-          chunk_index: 0,
-          chunk_text: 'A chunk that is embedded only in the dynamic column.',
-          chunk_source: 'compiled_truth',
-          embedding: new Float32Array(2560).fill(0.25),
-        },
-      ], { embeddingColumn: descriptor });
-
-      // Engine-level contrast: legacy predicate still sees the row as stale;
-      // the alt-column predicate does not.
-      expect(await local.countStaleChunks()).toBe(1);
-      expect(await local.countStaleChunks({ embeddingColumn: descriptor })).toBe(0);
-      // sumStaleChunkChars feeds the sync cost gate — same predicate contract.
-      expect(await local.sumStaleChunkChars()).toBeGreaterThan(0);
-      expect(await local.sumStaleChunkChars({ embeddingColumn: descriptor })).toBe(0);
-      expect(await local.listStaleChunks({ embeddingColumn: descriptor, batchSize: 100 })).toHaveLength(0);
-      expect(await local.listStaleChunks({ batchSize: 100 })).toHaveLength(1);
-
-      // Boundary-level: `embed --stale --dry-run` resolves the write column
-      // from merged config + gateway and reports NOTHING to embed. Without
-      // the fix this reports 1 (perpetual re-embed loop).
-      const result = await runEmbedCore(local, { stale: true, dryRun: true });
-      expect(result.would_embed).toBe(0);
-    } finally {
-      await local.disconnect();
-      if (previousHome === undefined) delete process.env.GBRAIN_HOME;
-      else process.env.GBRAIN_HOME = previousHome;
-      resetGateway();
-    }
-  });
-});
@@ -224,54 +224,4 @@ if (!dbUrl) {
      await engine.executeRaw(`UPDATE content_chunks SET embedding_voyage = '${v}'::vector WHERE id = ${dogId}`);
    });
  });
-
-  describe('Postgres: upsertChunks write-side ResolvedColumn descriptor (#1262)', () => {
-    const descriptor: ResolvedColumn = {
-      name: 'embedding_ze',
-      type: 'halfvec',
-      dimensions: 2560,
-      embeddingModel: 'zeroentropyai:zembed-1',
-    };
-
-    test('halfvec descriptor writes the text embedding to the alternate column, not legacy embedding', async () => {
-      await engine.putPage('docs/write-alt-postgres', {
-        type: 'concept',
-        title: 'Write alt column Postgres',
-        compiled_truth: 'Postgres write-side alternate embedding column test.',
-      });
-      await engine.upsertChunks('docs/write-alt-postgres', [
-        {
-          chunk_index: 0,
-          chunk_text: 'Postgres write-side alternate embedding column test.',
-          chunk_source: 'compiled_truth',
-          embedding: new Float32Array(2560).fill(0.25),
-        },
-      ], { embeddingColumn: descriptor });
-
-      const rows = await engine.executeRaw<{
-        has_default: boolean;
-        has_ze: boolean;
-      }>(
-        `SELECT embedding IS NOT NULL AS has_default,
-                embedding_ze IS NOT NULL AS has_ze
-           FROM content_chunks cc
-           JOIN pages p ON p.id = cc.page_id
-          WHERE p.slug = 'docs/write-alt-postgres'`,
-      );
-      expect(rows.length).toBe(1);
-      expect(rows[0].has_default).toBe(false);
-      expect(rows[0].has_ze).toBe(true);
-    }, 30_000);
-
-    test('stale scan follows the write-side column (count + list parity with the write target)', async () => {
-      // Legacy predicate: cat/dog/write-alt rows all have embedding NULL.
-      expect(await engine.countStaleChunks()).toBeGreaterThan(0);
-      // Alt-column predicate: every chunk has embedding_ze populated.
-      expect(await engine.countStaleChunks({ embeddingColumn: descriptor })).toBe(0);
-      expect(await engine.listStaleChunks({ embeddingColumn: descriptor, batchSize: 100 })).toHaveLength(0);
-      expect((await engine.listStaleChunks({ batchSize: 100 })).length).toBeGreaterThan(0);
-      // updated_desc arm uses the same predicate.
-      expect(await engine.listStaleChunks({ embeddingColumn: descriptor, orderBy: 'updated_desc', batchSize: 100 })).toHaveLength(0);
-    }, 30_000);
-  });
 }
@@ -43,26 +43,135 @@ describe('runExtractAtomsDrain (issue #1678)', () => {
    expect(batches).toBe(3);
  });

-  it('stops at the wallclock window with remaining > 0', async () => {
-    // SYNC stepping clock: now() #1 sets deadline (0+100=100); the while-check
-    // then sees 50, 50 (two batches), then 999999 → past deadline → stop.
-    const times = [0, 50, 50, 999_999];
-    let ti = 0;
-    const now = () => times[Math.min(ti++, times.length - 1)];
+  it('stops at the wallclock window; remaining is unknown (no post-window count)', async () => {
+    // Each batch consumes 60ms of the 100ms window: two batches fit, the
+    // third boundary check sees 120 ≥ 100 and stops. #2750: after the window
+    // elapses the final countRemaining is SKIPPED (it would overrun the
+    // window), so remaining reports null.
+    let now = 0;
    const result = await runExtractAtomsDrain(
      {
        withLock: passThroughLock,
        countRemaining: async () => 5, // never drains
-        runBatch: async () => ({ extracted: 1, skipped: 0 }),
-        now,
+        runBatch: async () => {
+          now += 60;
+          return { extracted: 1, skipped: 0 };
+        },
+        now: () => now,
      },
      { windowMs: 100 },
    );
    expect(result.stopped).toBe('window');
-    expect(result.remaining).toBe(5);
+    expect(result.remaining).toBeNull();
    expect(result.batches).toBe(2);
  });

+  it('passes one drain-level deadline signal into count and batch', async () => {
+    const seen: AbortSignal[] = [];
+    const controller = new AbortController();
+    let now = 0;
+    const result = await runExtractAtomsDrain(
+      {
+        withLock: passThroughLock,
+        countRemaining: async (signal) => {
+          seen.push(signal);
+          return 5;
+        },
+        runBatch: async (signal) => {
+          seen.push(signal);
+          now = 100;
+          return { extracted: 1, skipped: 0 };
+        },
+        now: () => now,
+      },
+      { windowMs: 100, abortSignal: controller.signal },
+    );
+    expect(result.stopped).toBe('window');
+    expect(result.batches).toBe(1);
+    expect(seen.length).toBe(2);
+    expect(seen[0]).toBe(seen[1]);
+    // Combined (timeout + external) signal, not the raw external one.
+    expect(seen[0]).not.toBe(controller.signal);
+  });
+
+  it('aborts a hung backlog count at the window deadline and releases the lock', async () => {
+    let released = false;
+    const result = await runExtractAtomsDrain(
+      {
+        withLock: async (work) => {
+          try { return await work(); }
+          finally { released = true; }
+        },
+        // Hangs until the drain's real-time deadline signal fires (10ms).
+        countRemaining: (signal) => new Promise((_resolve, reject) => {
+          signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+        }),
+        runBatch: async () => ({ extracted: 0, skipped: 0 }),
+        now: () => 0, // injected clock never advances — the SIGNAL must save us
+      },
+      { windowMs: 10 },
+    );
+    expect(result.stopped).toBe('window');
+    expect(result.remaining).toBeNull();
+    expect(released).toBe(true);
+  });
+
+  it('rethrows external cancellation after releasing the lock', async () => {
+    const controller = new AbortController();
+    let released = false;
+    const pending = runExtractAtomsDrain(
+      {
+        withLock: async (work) => {
+          try { return await work(); }
+          finally { released = true; }
+        },
+        countRemaining: (signal) => new Promise((_resolve, reject) => {
+          signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+        }),
+        runBatch: async () => ({ extracted: 0, skipped: 0 }),
+        now: () => 0,
+      },
+      { windowMs: 1_000_000, abortSignal: controller.signal },
+    );
+    controller.abort(new DOMException('worker timeout', 'AbortError'));
+    await expect(pending).rejects.toThrow('worker timeout');
+    expect(released).toBe(true);
+  });
+
+  it('classifies a deadline-exhausted zero-progress batch as window, not no_progress', async () => {
+    let now = 0;
+    const result = await runExtractAtomsDrain(
+      {
+        withLock: passThroughLock,
+        countRemaining: async () => 5,
+        runBatch: async () => {
+          now = 100; // batch consumed the whole window and returned nothing
+          return { extracted: 0, skipped: 0 };
+        },
+        now: () => now,
+      },
+      { windowMs: 100 },
+    );
+    expect(result.stopped).toBe('window');
+    expect(result.batches).toBe(1);
+  });
+
+  it('bounds a hung lock acquisition with the drain deadline signal', async () => {
+    const started = Date.now();
+    await expect(runExtractAtomsDrain(
+      {
+        withLock: (_work, signal) => new Promise((_resolve, reject) => {
+          signal.addEventListener('abort', () => reject(signal.reason), { once: true });
+        }),
+        countRemaining: async () => 1,
+        runBatch: async () => ({ extracted: 0, skipped: 0 }),
+        now: Date.now,
+      },
+      { windowMs: 10 },
+    )).rejects.toThrow();
+    expect(Date.now() - started).toBeLessThan(1_000);
+  });
+
  it('stops on a zero-progress batch (no hot loop)', async () => {
    let batches = 0;
    const result = await runExtractAtomsDrain(
@@ -133,4 +242,17 @@ describe('shared wiring helper holds the cycle lock (5A)', () => {
    expect(src).toContain('cycleLockIdFor(opts.sourceId)');
    expect(src).toContain('withRefreshingLock(engine, lockId');
  });
+
+  // #2750: the deadline signal must reach the phase, the backlog count, AND
+  // the lock wrapper — and the transcript path (brainDir) must stay wired
+  // exactly as the routine callers expect (PR #2752 takeover reverted its
+  // unsanctioned transcript-suppression scope change).
+  it('threads the drain deadline signal through phase, count, and lock', () => {
+    const jobsSrc = readFileSync(join(import.meta.dir, '../src/commands/jobs.ts'), 'utf8');
+    expect(src).toContain('abortSignal: signal');
+    expect(src).toContain('countExtractAtomsBacklog(engine, extractionSourceId, signal)');
+    expect(src).toContain('brainDir: opts.brainDir');
+    expect(src).not.toContain('_transcripts');
+    expect(jobsSrc).toContain('abortSignal: job.signal');
+  });
 });
@@ -24,15 +24,18 @@ import type { BrainEngine } from '../../src/core/engine.ts';
 // Mock engine: healthCheck() calls engine.executeRaw; return empty rows so
 // the query path exercises without needing Postgres.
 //
-// #1849: start() now acquires the queue-scoped DB singleton lock via
-// tryAcquireDbLock, which uses the postgres `sql` tagged-template escape hatch.
-// The stub returns a single row from every call so acquire succeeds (length 1
-// → acquired) and refresh/release are no-ops. Each spawned runner is a fresh
-// process, so there's no cross-test lock state to clean up.
+// #1849: start() acquires the queue-scoped DB singleton lock via
+// tryAcquireDbLock. #2750 routed the acquire upsert through engine.executeRaw
+// (signal-boundable) and release through engine.executeRawDirect, so the
+// stub returns a single row from the lock upsert (length 1 → acquired) and
+// empty rows everywhere else. Each spawned runner is a fresh process, so
+// there's no cross-test lock state to clean up.
 const sqlStub = (..._args: unknown[]) => Promise.resolve([{ id: 'supervisor-lock' }]);
 const mockEngine: Partial<BrainEngine> = {
  kind: 'postgres' as const,
-  executeRaw: async () => [],
+  executeRaw: async (query: string) =>
+    query.includes('gbrain_cycle_locks') ? [{ id: 'supervisor-lock' }] : [],
+  executeRawDirect: async () => [],
  sql: sqlStub,
 } as unknown as BrainEngine;

@@ -98,14 +98,39 @@ describe('PostgresEngine.executeRawDirect — routing decision (PR #1816)', () =
  });

  test('already-aborted signal short-circuits with AbortError before routing the query', async () => {
-    const readConn = fakeSql('read');
+    let unsafeCalls = 0;
+    let ddlCalls = 0;
+    const readConn: FakeSql = { unsafe: async () => { unsafeCalls++; return []; } };
    const directConn = fakeSql('direct');
    const engine = makeEngine({ dualPoolActive: true, readConn, directConn });
+    const e = engine as unknown as { connectionManager: { ddl: () => Promise<FakeSql> } };
+    e.connectionManager.ddl = async () => { ddlCalls++; return directConn; };

    const ac = new AbortController();
    ac.abort();
    await expect(
      engine.executeRawDirect('UPDATE minion_jobs SET x=1', [], { signal: ac.signal }),
    ).rejects.toThrow(/abort/i);
+    // #2750: short-circuits BEFORE pool routing — no ddl(), no unsafe().
+    expect(ddlCalls).toBe(0);
+    expect(unsafeCalls).toBe(0);
+  });
+
+  test('#2750: signal bounds a stalled direct-pool acquisition before unsafe starts', async () => {
+    let unsafeCalls = 0;
+    const readConn: FakeSql = { unsafe: async () => { unsafeCalls++; return []; } };
+    const directConn = fakeSql('direct');
+    const engine = makeEngine({ dualPoolActive: true, readConn, directConn });
+    const e = engine as unknown as { connectionManager: { ddl: () => Promise<FakeSql> } };
+    e.connectionManager.ddl = () => new Promise<FakeSql>(() => {}); // pooler exhausted: never resolves
+
+    const started = Date.now();
+    await expect(engine.executeRawDirect(
+      'DELETE FROM gbrain_cycle_locks',
+      [],
+      { signal: AbortSignal.timeout(10) },
+    )).rejects.toThrow(/abort/i);
+    expect(Date.now() - started).toBeLessThan(1_000);
+    expect(unsafeCalls).toBe(0);
  });
 });
@@ -13,10 +13,9 @@
 *     throw on unknown string.
 */

-import { describe, test, expect, afterAll, afterEach } from 'bun:test';
+import { describe, test, expect } from 'bun:test';
 import {
  resolveEmbeddingColumn,
-  resolveWriteColumn,
  getEmbeddingColumnRegistry,
  buildVectorCastFragment,
  quoteIdentifier,
@@ -35,28 +34,6 @@ import {
 } from '../../src/core/search/embedding-column.ts';
 import type { GBrainConfig } from '../../src/core/config.ts';
 import type { ResolvedColumn } from '../../src/core/types.ts';
-import { configureGateway, resetGateway } from '../../src/core/ai/gateway.ts';
-
-/**
- * Teardown: reset AND re-apply the legacy preload config
- * (test/helpers/legacy-embedding-preload.ts). A bare resetGateway() would
- * leave the slot empty for the NEXT file's beforeAll (the preload's
- * per-test beforeEach only fires before tests, not before beforeAll), which
- * would make sibling PGLite fixtures initSchema at the 1280 default instead
- * of the legacy 1536 their seed vectors assume.
- */
-function restorePreloadGateway() {
-  resetGateway();
-  configureGateway({
-    embedding_model: 'openai:text-embedding-3-large',
-    embedding_dimensions: 1536,
-    env: { ...process.env },
-  });
-}
-
-afterAll(() => {
-  restorePreloadGateway();
-});

 function cfg(overrides: Partial<GBrainConfig> = {}): GBrainConfig {
  return { engine: 'pglite', ...overrides };
@@ -545,89 +522,3 @@ describe('codex /ship #4 — isCacheSafe (embedding-space-based skip)', () => {
    expect(isCacheSafe(r, cfg())).toBe(true);
  });
 });
-
-describe('resolveWriteColumn — write-side boundary resolution (#1262)', () => {
-  afterEach(() => {
-    restorePreloadGateway();
-  });
-
-  test('no registry / empty registry returns undefined (legacy single-column brain)', () => {
-    expect(resolveWriteColumn(cfg())).toBeUndefined();
-    expect(resolveWriteColumn(cfg({ embedding_columns: {} }))).toBeUndefined();
-  });
-
-  test('provider match via cfg.embedding_model returns the descriptor', () => {
-    const r = resolveWriteColumn(cfg({
-      embedding_model: 'voyage:voyage-3-large',
-      embedding_dimensions: 1024,
-      embedding_columns: {
-        embedding_voyage: { provider: 'voyage:voyage-3-large', dimensions: 1024, type: 'vector' },
-      },
-    }));
-    expect(r).toEqual({
-      name: 'embedding_voyage',
-      type: 'vector',
-      dimensions: 1024,
-      embeddingModel: 'voyage:voyage-3-large',
-    });
-  });
-
-  test('provider match via gateway state (cfg.embedding_model unset) returns descriptor', () => {
-    configureGateway({
-      embedding_model: 'zeroentropyai:zembed-1',
-      embedding_dimensions: 2560,
-      env: {},
-    });
-    const r = resolveWriteColumn(cfg({
-      embedding_columns: {
-        embedding_ze: { provider: 'zeroentropyai:zembed-1', dimensions: 2560, type: 'halfvec' },
-      },
-    }));
-    expect(r).toEqual({
-      name: 'embedding_ze',
-      type: 'halfvec',
-      dimensions: 2560,
-      embeddingModel: 'zeroentropyai:zembed-1',
-    });
-  });
-
-  test('no provider match returns undefined instead of guessing a column', () => {
-    configureGateway({
-      embedding_model: 'zeroentropyai:zembed-1',
-      embedding_dimensions: 2560,
-      env: {},
-    });
-    const r = resolveWriteColumn(cfg({
-      embedding_columns: {
-        embedding_voyage: { provider: 'voyage:voyage-3-large', dimensions: 1024, type: 'vector' },
-      },
-    }));
-    expect(r).toBeUndefined();
-  });
-
-  test('only USER-declared columns are consulted — multimodal builtin never captures text writes', () => {
-    // Current model equals the embedding_image BUILTIN's provider; a registry
-    // walk that consulted builtins would misroute text writes into the image
-    // column. resolveWriteColumn must return undefined here.
-    configureGateway({
-      embedding_model: 'voyage:voyage-multimodal-3',
-      embedding_dimensions: 1024,
-      env: {},
-    });
-    const r = resolveWriteColumn(cfg({
-      embedding_columns: {
-        embedding_other: { provider: 'openai:text-embedding-3-large', dimensions: 1536, type: 'vector' },
-      },
-    }));
-    expect(r).toBeUndefined();
-  });
-
-  test('malformed registry entry throws loud (same validation as the read side)', () => {
-    expect(() => resolveWriteColumn(cfg({
-      embedding_model: 'voyage:voyage-3-large',
-      embedding_columns: {
-        'bad"col': { provider: 'voyage:voyage-3-large', dimensions: 1024, type: 'vector' },
-      } as never,
-    }))).toThrow(EmbeddingColumnConfigError);
-  });
-});