fix(facts): seed phase B backfill row_nums past the DB max (#2044 class)

phaseBFenceFacts appended at fence-max+1 with no DB seeding, so on a
diverged page (stamped DB rows + missing fence at the resolved path —
the same divergence class the writeFactsToFence fix handles) the
post-rename UPDATE hit idx_facts_fence_key, the page failed, and the
already-renamed fence disagreed with the DB while the row_num-NULL
backlog persisted. Seed appended row_nums from MAX(fence max, DB max)
per (source_id, slug), mirroring the fence-write fix. Repro test added.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

src/cli.ts

@@ -54,7 +54,7 @@ export function bigintToStringReplacer(_key: string, value: unknown): unknown {
 }
 
 // CLI-only commands that bypass the operation layer
-export const CLI_ONLY = new Set(['init', 'reinit-pglite', 'upgrade', 'post-upgrade', 'check-update', 'integrations', 'publish', 'check-backlinks', 'lint', 'report', 'import', 'export', 'files', 'embed', 'serve', 'call', 'config', 'doctor', 'migrate', 'eval', 'sync', 'extract', 'extract-conversation-facts', 'enrich', 'features', 'autopilot', 'graph-query', 'jobs', 'agent', 'apply-migrations', 'skillpack-check', 'skillpack', 'resolvers', 'integrity', 'repair-jsonb', 'orphans', 'sources', 'mounts', 'dream', 'check-resolvable', 'routing-eval', 'skillify', 'smoke-test', 'providers', 'storage', 'repos', 'code-def', 'code-refs', 'reindex', 'reindex-code', 'reindex-frontmatter', 'code-callers', 'code-callees', 'reconcile-links', 'frontmatter', 'auth', 'friction', 'claw-test', 'book-mirror', 'takes', 'think', 'salience', 'anomalies', 'calibration', 'transcripts', 'models', 'remote', 'recall', 'forget', 'edges-backfill', 'cache', 'ze-switch', 'founder', 'brainstorm', 'lsd', 'schema', 'capture', 'onboard', 'conversation-parser', 'status', 'connect', 'skillopt', 'quarantine', 'self-upgrade', 'advisor', 'watch', 'reindex-search-vector']);
+export const CLI_ONLY = new Set(['init', 'reinit-pglite', 'upgrade', 'post-upgrade', 'check-update', 'integrations', 'publish', 'check-backlinks', 'lint', 'report', 'import', 'export', 'files', 'embed', 'serve', 'call', 'config', 'doctor', 'migrate', 'eval', 'sync', 'extract', 'extract-conversation-facts', 'enrich', 'features', 'autopilot', 'graph-query', 'jobs', 'agent', 'apply-migrations', 'skillpack-check', 'skillpack', 'resolvers', 'integrity', 'repair-jsonb', 'orphans', 'sources', 'mounts', 'dream', 'check-resolvable', 'routing-eval', 'skillify', 'smoke-test', 'providers', 'storage', 'repos', 'code-def', 'code-refs', 'reindex', 'reindex-code', 'reindex-frontmatter', 'code-callers', 'code-callees', 'reconcile-links', 'frontmatter', 'auth', 'friction', 'claw-test', 'book-mirror', 'takes', 'think', 'salience', 'anomalies', 'calibration', 'transcripts', 'models', 'remote', 'recall', 'forget', 'edges-backfill', 'facts', 'cache', 'ze-switch', 'founder', 'brainstorm', 'lsd', 'schema', 'capture', 'onboard', 'conversation-parser', 'status', 'connect', 'skillopt', 'quarantine', 'self-upgrade', 'advisor', 'watch', 'reindex-search-vector']);
 // CLI-only commands whose handlers print their own --help text. These are
 // excluded from the generic short-circuit so detailed per-command and
 // per-subcommand usage stays reachable.
@@ -991,6 +991,8 @@ const THIN_CLIENT_REFUSED_COMMANDS = new Set([
   // hint pointing at the routable MCP tools; per-subcommand splits are
   // a v0.31.x follow-up TODO.
   'takes', 'sources',
+  // #1867: fence-backfill edits local .md fences + stamps the local DB.
+  'facts',
   // v0.32 thin-client routing audit (Codex round 2 findings #2, #4):
   // - `pages` purge-deleted is admin+localOnly (operations.ts:856-864)
   // - `files` list / file_url MCP ops are localOnly (operations.ts:1769-1879)
@@ -1026,6 +1028,7 @@ const THIN_CLIENT_REFUSE_HINTS: Record<string, string> = {
   migrate: "migrate runs on the host's local engine. Run on the host machine.",
   'apply-migrations': 'schema migrations run on the host. SSH and run there.',
   'repair-jsonb': 'repair-jsonb operates on the local DB only.',
+  facts: 'facts fence-backfill edits local entity-page fences. Run on the host machine.',
   integrity: 'integrity scans local files. Run on the host machine.',
   serve: 'serve starts a server. Run on the host, not the thin client.',
   dream: 'dream runs the autopilot cycle on the host. `gbrain remote ping` queues one. (Native `gbrain dream` thin-client routing planned for v0.31.2.)',
@@ -1855,6 +1858,13 @@ async function handleCliOnly(command: string, args: string[]) {
         await runEdgesBackfill(engine, args);
         break;
       }
+      case 'facts': {
+        // #1867 — re-runnable fence-backfill for row_num-NULL legacy fact
+        // rows (idempotent v0_32_2 phase B, exposed as an operator command).
+        const { runFactsCommand } = await import('./commands/facts.ts');
+        await runFactsCommand(engine, args);
+        break;
+      }
       case 'whoknows': {
         // v0.33 (Issue #?): expertise + relationship-proximity routing.
         // MCP op `find_experts` (read-scoped) backs the same code path; CLI
@@ -2335,6 +2345,7 @@ TOOLS
   check-backlinks <check|fix> [dir]  Find/fix missing back-links across brain
   lint <dir|file> [--fix]            Catch LLM artifacts, placeholder dates, bad frontmatter
   orphans [--json] [--count]         Find pages with no inbound wikilinks
+  facts fence-backfill [--dry-run]   Fence legacy fact rows (row_num NULL) onto entity pages
   salience [--days N] [--kind P]     v0.29: pages ranked by emotional + activity salience
   anomalies [--since D] [--sigma N]  v0.29: cohort-based statistical anomalies (tag, type)
   transcripts recent [--days N]      v0.29: recent raw .txt transcripts (local-only)

src/commands/doctor.ts

@@ -820,10 +820,6 @@ export async function doctorReportRemote(engine: BrainEngine): Promise<DoctorRep
   // v0.42.x (#1794, 4A): pool-budget nudge when GBRAIN_MAX_CONNECTIONS is set.
   checks.push(await checkPoolBudget(engine));
 
-  // #2552: warn when an explicit embed-concurrency override fans out against
-  // a local single-slot embedding endpoint (silent backfill starvation).
-  checks.push(await checkEmbedConcurrency());
-
   // v0.42.7 (#1696): link-extraction lag. Strictly SQL (single indexed COUNT),
   // safe on the thin-client/remote path — remote operators on checkout-less
   // Postgres brains are exactly who can't otherwise see the extraction backlog.
@@ -3819,61 +3815,6 @@ export function computePoolBudgetCheck(
   };
 }
 
-/**
- * #2552: warn when an explicit GBRAIN_EMBED_CONCURRENCY override fans out
- * against a local single-slot embedding endpoint (Ollama / llama-server /
- * localhost base URL). Requests serialize on the one loaded model, so N
- * parallel pages multiply latency xN and can exceed the fetch timeout with
- * no surfaced error — the backfill silently starves. (When the env var is
- * unset, embed auto-caps at LOCAL_EMBED_CONCURRENCY_CAP and this check
- * reports ok.) Pure; exported for tests.
- */
-export function computeEmbedConcurrencyCheck(
-  isLocalEndpoint: boolean,
-  envValue: string | undefined,
-  localCap: number,
-): Check {
-  const name = 'embed_concurrency';
-  if (!isLocalEndpoint) {
-    return { name, status: 'ok', message: 'Embedding endpoint is not a local inference server — cloud concurrency defaults apply.' };
-  }
-  const parsed = envValue ? parseInt(envValue, 10) : NaN;
-  if (envValue && Number.isFinite(parsed) && parsed > localCap) {
-    return {
-      name,
-      status: 'warn',
-      message:
-        `GBRAIN_EMBED_CONCURRENCY=${parsed} against a local embedding endpoint. ` +
-        `Local inference servers serialize requests, so ${parsed} parallel pages multiply ` +
-        `latency x${parsed} and can exceed the fetch timeout — the embed backfill stalls ` +
-        `with no error. Unset GBRAIN_EMBED_CONCURRENCY (auto-caps at ${localCap}) or set it <= ${localCap}.`,
-    };
-  }
-  return {
-    name,
-    status: 'ok',
-    message: `Local embedding endpoint detected; embed concurrency capped at ${envValue ? parsed : localCap}.`,
-  };
-}
-
-/** Thin gateway/env wrapper over `computeEmbedConcurrencyCheck`. */
-export async function checkEmbedConcurrency(): Promise<Check> {
-  try {
-    const { isLocalEmbeddingEndpoint, LOCAL_EMBED_CONCURRENCY_CAP } = await import('../core/ai/gateway.ts');
-    return computeEmbedConcurrencyCheck(
-      isLocalEmbeddingEndpoint(),
-      process.env.GBRAIN_EMBED_CONCURRENCY,
-      LOCAL_EMBED_CONCURRENCY_CAP,
-    );
-  } catch (err) {
-    return {
-      name: 'embed_concurrency',
-      status: 'ok',
-      message: `Skipped (${err instanceof Error ? err.message : String(err)})`,
-    };
-  }
-}
-
 /** Thin env/engine wrapper over `computePoolBudgetCheck`. */
 export async function checkPoolBudget(_engine: BrainEngine): Promise<Check> {
   try {

src/commands/embed.ts

@@ -1,6 +1,5 @@
 import type { BrainEngine } from '../core/engine.ts';
 import { embedBatch, currentEmbeddingSignature } from '../core/embedding.ts';
-import { isLocalEmbeddingEndpoint, LOCAL_EMBED_CONCURRENCY_CAP } from '../core/ai/gateway.ts';
 import type { ChunkInput } from '../core/types.ts';
 import { chunkText } from '../core/chunkers/recursive.ts';
 import { createProgress, type ProgressReporter } from '../core/progress.ts';
@@ -177,31 +176,6 @@ export class EmbeddingDimMismatchError extends Error {
   }
 }
 
-/**
- * #2552: resolve the bulk-embed worker count. Env override or the
- * cloud-tuned default of 20 — but when the operator did NOT set
- * GBRAIN_EMBED_CONCURRENCY and the embedding endpoint is a local inference
- * server (Ollama / llama-server / localhost base URL), cap at
- * LOCAL_EMBED_CONCURRENCY_CAP: 20 parallel pages against a single-slot
- * server serialize on the one loaded model, multiply latency x20 past the
- * fetch timeout, and starve the backfill with no surfaced error. An
- * explicit env value always wins (`gbrain doctor` warns instead).
- * Pacing only ever LOWERS concurrency (Codex P2).
- */
-export function resolveEmbedConcurrency(paceMaxConcurrency?: number): number {
-  const envSet = !!process.env.GBRAIN_EMBED_CONCURRENCY;
-  const base = parseInt(process.env.GBRAIN_EMBED_CONCURRENCY || '20', 10);
-  let resolved = base;
-  if (!envSet && isLocalEmbeddingEndpoint() && base > LOCAL_EMBED_CONCURRENCY_CAP) {
-    resolved = LOCAL_EMBED_CONCURRENCY_CAP;
-    serr(
-      `[embed] local embedding endpoint detected — capping concurrency at ` +
-      `${LOCAL_EMBED_CONCURRENCY_CAP} (set GBRAIN_EMBED_CONCURRENCY to override)`,
-    );
-  }
-  return paceMaxConcurrency ? Math.min(resolved, paceMaxConcurrency) : resolved;
-}
-
 /**
  * Pre-flight check: read the actual schema column dim and compare to the
  * gateway's resolved dim. Throws `EmbeddingDimMismatchError` on mismatch
@@ -703,8 +677,10 @@ async function embedAll(
   // Paced runs lower this to the resolved cap (the real lever vs pooler-slot
   // starvation); unpaced keeps the env/default 20. Codex P2: only ever LOWER —
   // never raise above an operator's existing env cap.
-  // #2552: local endpoints auto-cap — see resolveEmbedConcurrency.
-  const CONCURRENCY = resolveEmbedConcurrency(staleOpts?.paceMaxConcurrency);
+  const BASE_CONCURRENCY = parseInt(process.env.GBRAIN_EMBED_CONCURRENCY || '20', 10);
+  const CONCURRENCY = staleOpts?.paceMaxConcurrency
+    ? Math.min(BASE_CONCURRENCY, staleOpts.paceMaxConcurrency)
+    : BASE_CONCURRENCY;
 
   async function embedOnePage(page: typeof pages[number]) {
     // #1737: bail before doing any work for this page if the run was aborted.
@@ -879,8 +855,10 @@ async function embedAllStale(
   // Paced runs lower concurrency to the resolved cap (E-1: worker count IS the
   // lever on this single pool, no separate permit). Codex P2: pacing only ever
   // LOWERS concurrency — never raise above an operator's existing env cap.
-  // #2552: local endpoints auto-cap — see resolveEmbedConcurrency.
-  const CONCURRENCY = resolveEmbedConcurrency(staleOpts?.paceMaxConcurrency);
+  const BASE_CONCURRENCY = parseInt(process.env.GBRAIN_EMBED_CONCURRENCY || '20', 10);
+  const CONCURRENCY = staleOpts?.paceMaxConcurrency
+    ? Math.min(BASE_CONCURRENCY, staleOpts.paceMaxConcurrency)
+    : BASE_CONCURRENCY;
   const pacer = staleOpts?.pacer ?? createNoopPacer();
 
   // D3 + D3a + D8: wall-clock budget. 30 min default; env override.

src/commands/facts.ts

@@ -0,0 +1,48 @@
+/**
+ * gbrain facts — fact-store maintenance surface (#1867).
+ *
+ * `fence-backfill` re-runs the v0_32_2 fence-backfill phase on demand.
+ * Remote `extract_facts` deposits that predate the fence-write backstop
+ * (and any legacy DB-only insert) leave `row_num IS NULL` rows that the
+ * cycle extract_facts guard refuses to reconcile past — previously the
+ * only remedy was the one-shot v0_32_2 migration, which the ledger marks
+ * complete and never re-runs. The phase is idempotent (only touches
+ * `row_num IS NULL` rows), so exposing it as a command is safe to re-run
+ * any time the backlog reappears.
+ */
+import type { BrainEngine } from '../core/engine.ts';
+import { setCliExitVerdict } from '../core/cli-force-exit.ts';
+import { phaseBFenceFacts } from './migrations/v0_32_2.ts';
+
+function printHelp(): void {
+  process.stderr.write(
+    `Usage: gbrain facts fence-backfill [--dry-run]\n\n` +
+      `Fence-backfill: appends every legacy fact row (row_num IS NULL) to its\n` +
+      `entity page's \`## Facts\` fence and stamps row_num + source_markdown_slug\n` +
+      `back onto the DB row. Idempotent — re-runs only pick up rows still\n` +
+      `missing a fence assignment. Clears the backlog that makes the cycle's\n` +
+      `extract_facts phase skip fence→DB reconciliation.\n\n` +
+      `  --dry-run   report what would be fenced; no FS or DB writes\n`,
+  );
+}
+
+export async function runFactsCommand(engine: BrainEngine, args: string[]): Promise<void> {
+  const sub = args[0];
+  if (!sub || sub === '--help' || sub === '-h') {
+    printHelp();
+    return;
+  }
+  if (sub !== 'fence-backfill') {
+    process.stderr.write(`Unknown facts subcommand: ${sub}\n`);
+    printHelp();
+    setCliExitVerdict(1);
+    return;
+  }
+
+  const dryRun = args.includes('--dry-run');
+  const result = await phaseBFenceFacts(engine, { dryRun });
+  process.stderr.write(
+    `fence-backfill: ${result.status}${result.detail ? ` — ${result.detail}` : ''}\n`,
+  );
+  if (result.status === 'failed') setCliExitVerdict(1);
+}

src/commands/migrations/v0_32_2.ts

@@ -39,6 +39,7 @@ import type { BrainEngine } from '../../core/engine.ts';
 import { loadConfig, toEngineConfig } from '../../core/config.ts';
 import { createEngine } from '../../core/engine-factory.ts';
 import { upsertFactRow, parseFactsFence } from '../../core/facts-fence.ts';
+import { resolvePageFilePath } from '../../core/markdown.ts';
 
 let testEngineOverride: BrainEngine | null = null;
 export function __setTestEngineOverride(engine: BrainEngine | null): void {
@@ -148,9 +149,16 @@ function isLocalPathDirty(localPath: string): boolean {
   }
 }
 
-async function phaseBFenceFacts(
+/**
+ * Exported (not just via `__testing`) because `gbrain facts fence-backfill`
+ * (#1867) re-runs this phase on demand: remote `extract_facts` deposits that
+ * predate the fence-write backstop leave row_num-NULL rows the cycle guard
+ * refuses to reconcile past. The phase is idempotent (only touches
+ * `row_num IS NULL` rows), so re-running is always safe.
+ */
+export async function phaseBFenceFacts(
   engine: BrainEngine | null,
-  opts: OrchestratorOpts,
+  opts: Pick<OrchestratorOpts, 'dryRun'>,
 ): Promise<OrchestratorPhaseResult> {
   if (opts.dryRun) {
     // Dry-run: report what WOULD happen without touching FS or DB.
@@ -238,7 +246,11 @@ async function phaseBFenceFacts(
     for (const [key, group] of groups) {
       const [sourceId, entitySlug] = key.split('\0');
       const localPath = localPathById.get(sourceId)!;
-      const filePath = join(localPath, `${entitySlug}.md`);
+      // resolvePageFilePath, NOT a bare join — non-default sources fence
+      // into `<local_path>/.sources/<id>/<slug>.md`, the same path the
+      // fence-write backstop and put_page write-through compute. A bare
+      // join here diverges fence and DB for non-default sources (#2044).
+      const filePath = resolvePageFilePath(localPath, entitySlug, sourceId);
       const tmpPath = `${filePath}.tmp`;
 
       try {
@@ -269,6 +281,21 @@ async function phaseBFenceFacts(
         const existingFence = parseFactsFence(body);
         const existingKeySet = new Set(existingFence.facts.map(f => `${f.claim}\0${f.source ?? ''}`));
 
+        // Seed appended row_nums from MAX(fence max, DB max) — same #2044
+        // divergence guard as writeFactsToFence. When the fence at the
+        // resolved path is missing/behind but the DB already holds stamped
+        // rows for this slug (legacy wrong-path fence writes), fence-max+1
+        // collides with idx_facts_fence_key on the post-rename UPDATE,
+        // failing the page and leaving fence and DB disagreeing.
+        const dbMaxRows = await engine.executeRaw<{ max: number | string | null }>(
+          `SELECT MAX(row_num) AS max FROM facts
+            WHERE source_id = $1 AND source_markdown_slug = $2`,
+          [sourceId, entitySlug],
+        );
+        const dbMaxRowNum = Number(dbMaxRows[0]?.max ?? 0) || 0;
+        const fenceMaxRowNum = existingFence.facts.reduce((m, f) => Math.max(m, f.rowNum), 0);
+        let nextRowNum = Math.max(fenceMaxRowNum, dbMaxRowNum) + 1;
+
         const assignments: Array<{ id: string; row_num: number }> = [];
         for (const row of group) {
           const key = `${row.fact}\0${row.source ?? ''}`;
@@ -291,6 +318,7 @@ async function phaseBFenceFacts(
                 .toISOString().slice(0, 10)
             : undefined;
           const { body: updated, rowNum } = upsertFactRow(body, {
+            rowNum:     nextRowNum++,
             claim:      row.fact,
             kind:       row.kind,
             confidence: row.confidence,
@@ -381,7 +409,7 @@ async function phaseCVerify(
     for (const g of groups) {
       const localPath = localPathById.get(g.source_id);
       if (!localPath) continue;
-      const filePath = join(localPath, `${g.source_markdown_slug}.md`);
+      const filePath = resolvePageFilePath(localPath, g.source_markdown_slug, g.source_id);
       if (!existsSync(filePath)) {
         mismatches.push(`${g.source_markdown_slug} (file missing)`);
         continue;

src/core/ai/gateway.ts

@@ -683,33 +683,6 @@ export function getEmbeddingDimensions(): number {
   return requireConfig().embedding_dimensions ?? DEFAULT_EMBEDDING_DIMENSIONS;
 }
 
-/**
- * #2552: cap for parallel bulk-embed workers against a local inference
- * server. A single-slot Ollama/llama-server serializes requests, so the
- * cloud-tuned 20-worker fan-out multiplies latency x20 and blows past the
- * fetch timeout with no surfaced error (the backfill silently starves).
- */
-export const LOCAL_EMBED_CONCURRENCY_CAP = 2;
-
-/**
- * #2552: true when the configured embedding model routes to a local
- * inference server — the `ollama` / `llama-server` recipes, or any recipe
- * whose base URL was explicitly pointed at localhost. Bulk callers use this
- * to pick CPU-safe concurrency defaults; `gbrain doctor` uses it to warn
- * about an explicit cloud-sized override. Fail-open: unconfigured or
- * unresolvable gateway → false (cloud behavior, the historical default).
- */
-export function isLocalEmbeddingEndpoint(): boolean {
-  try {
-    const { recipe } = resolveRecipe(getEmbeddingModel());
-    if (recipe.id === 'ollama' || recipe.id === 'llama-server') return true;
-    const base = requireConfig().base_urls?.[recipe.id] ?? '';
-    return /\/\/(localhost|127\.0\.0\.1|\[::1\])(:|\/|$)/i.test(base);
-  } catch {
-    return false;
-  }
-}
-
 /**
  * v0.28.11: returns the configured multimodal embedding model when set,
  * or undefined if the brain falls back to `embedding_model` for multimodal

src/core/ai/recipes/ollama.ts

@@ -29,17 +29,9 @@ export const ollama: Recipe = {
       trust_custom_dims: true, // #2271: local models carry varied native dims
       cost_per_1m_tokens_usd: 0,
       price_last_verified: '2026-04-20',
-      // #2552: Ollama's true batch capacity depends on the locally loaded
-      // model + OLLAMA_NUM_PARALLEL, but the previous `no_batch_cap: true`
-      // meant a whole page went out in ONE request — on a CPU-only box that
-      // multiplies latency past the fetch timeout and the backfill starves
-      // with no surfaced error. Ollama doesn't return a recognizable
-      // token-limit error either, so the recursive-halving safety net never
-      // fires; a conservative static pre-split cap is the only guard.
-      // 4096 tokens x 2 chars/token ~= 8K chars per request (code-dense
-      // pages run ~2 chars/token, not the tiktoken-ish 4).
-      max_batch_tokens: 4096,
-      chars_per_token: 2,
+      // Ollama's batch capacity depends on the locally loaded model + the
+      // OLLAMA_NUM_PARALLEL config; no static cap to declare. v0.32 (#779).
+      no_batch_cap: true,
     },
   },
   setup_hint: 'Install Ollama from https://ollama.ai, then `ollama pull nomic-embed-text` and `ollama serve`.',

src/core/cycle/extract-facts.ts

@@ -176,9 +176,11 @@ export async function runExtractFacts(
   if (legacyCount > 0) {
     result.guardTriggered = true;
     result.warnings.push(
-      `extract_facts: ${legacyCount} legacy v0.31 fact rows pending fence backfill. ` +
-      `Run \`gbrain apply-migrations --yes\` to complete v0_32_2 before this phase ` +
-      `can safely reconcile fence → DB.`,
+      `extract_facts: ${legacyCount} legacy fact rows pending fence backfill ` +
+      `(row_num IS NULL — v0.31 rows or remote extract_facts deposits that ` +
+      `predate the fence backstop). Run \`gbrain facts fence-backfill\` ` +
+      `(idempotent, re-runnable) before this phase can safely reconcile ` +
+      `fence → DB.`,
     );
     return result;
   }

src/core/doctor-categories.ts

@@ -141,7 +141,6 @@ export const OPS_CHECK_NAMES: ReadonlySet<string> = new Set([
   'pgbouncer_prepare',
   'pgvector',
   'pool_budget',
-  'embed_concurrency',
   'progressive_batch_audit_health',
   'queue_health',
   'reranker_health',

src/core/engine.ts

@@ -1739,13 +1739,14 @@ export interface BrainEngine {
    * single-row supersede flow because fence reconciliation is the canonical
    * source-of-truth direction, not the consolidator path.
    *
-   * Insertion is atomic per call: all rows commit in a single transaction
-   * or none commit (the transaction rolls back on any constraint
-   * violation, e.g. the v51 partial UNIQUE index on
-   * `(source_id, source_markdown_slug, row_num)`).
+   * Insertion runs in a single transaction. A collision on the v51
+   * partial UNIQUE index `(source_id, source_markdown_slug, row_num)`
+   * skips ONLY that row (ON CONFLICT DO NOTHING, #2044) — the rest of
+   * the batch still commits, so a redundant deposit against an
+   * already-indexed fence row is idempotent instead of a hard failure.
    *
-   * Returns the inserted ids in input-order so callers can correlate
-   * fence-row → DB-id without a separate lookup.
+   * Returns the inserted ids in input-order (colliding rows omitted) so
+   * callers can correlate fence-row → DB-id without a separate lookup.
    */
   insertFacts(
     rows: Array<NewFact & { row_num: number; source_markdown_slug: string }>,

src/core/facts/fence-write.ts

@@ -218,11 +218,27 @@ export async function writeFactsToFence(
       }
 
       // 2. Upsert each fact onto the fence in input order. row_num
-      //    monotonically increases (max-existing + 1 per call, append-only).
+      //    monotonically increases, append-only, seeded from the MAX of
+      //    the fence and the DB index (#2044): when fence and DB have
+      //    diverged (e.g. legacy writes that stamped DB rows against a
+      //    fence at a path this code no longer reads), fence-max+1 can
+      //    collide with an existing DB row_num, tripping
+      //    idx_facts_fence_key and rolling back the whole batch.
+      const dbMaxRows = await engine.executeRaw<{ max: number | string | null }>(
+        `SELECT MAX(row_num) AS max FROM facts
+          WHERE source_id = $1 AND source_markdown_slug = $2`,
+        [target.sourceId, target.slug],
+      );
+      const dbMaxRowNum = Number(dbMaxRows[0]?.max ?? 0) || 0;
+      const fenceMaxRowNum = parseFactsFence(body).facts
+        .reduce((m, f) => Math.max(m, f.rowNum), 0);
+      let nextRowNum = Math.max(fenceMaxRowNum, dbMaxRowNum) + 1;
+
       const assignedRowNums: number[] = [];
       for (const f of facts) {
         const validFromStr = (f.validFrom ?? new Date()).toISOString().slice(0, 10);
         const { body: updated, rowNum } = upsertFactRow(body, {
+          rowNum:      nextRowNum++,
           claim:       f.fact,
           kind:        (f.kind ?? 'fact') as 'fact' | 'event' | 'preference' | 'commitment' | 'belief',
           confidence:  f.confidence ?? 1.0,

src/core/pglite-engine.ts

@@ -56,7 +56,7 @@ import { GBrainError, PAGE_SORT_SQL, ENRICH_ORDER_SQL } from './types.ts';
 import { finalizeLastSeen } from './chronicle/last-seen.ts';
 import { computeAnomaliesFromBuckets } from './cycle/anomaly.ts';
 import { resolveBoostMap, resolveHardExcludes } from './search/source-boost.ts';
-import { buildSourceFactorCase, buildHardExcludeClause, buildVisibilityClause, buildRecencyComponentSql, buildBestPerPagePoolCte, buildOrFallbackWebsearchQuery, buildWebsearchQueryExpr } from './search/sql-ranking.ts';
+import { buildSourceFactorCase, buildHardExcludeClause, buildVisibilityClause, buildRecencyComponentSql, buildBestPerPagePoolCte, buildOrFallbackWebsearchQuery } from './search/sql-ranking.ts';
 import {
   normalizeEngineColumn,
   buildVectorCastFragment,
@@ -1591,8 +1591,6 @@ export class PGLiteEngine implements BrainEngine {
     }
 
     // v0.20.0 Cathedral II Layer 10 C1/C2: language + symbol-kind filters.
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query, innerLimit, limit, offset];
     let extraFilter = '';
     if (opts?.language) {
@@ -1632,7 +1630,6 @@ export class PGLiteEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
     const keywordSql =
       `WITH ranked AS (
@@ -1640,14 +1637,14 @@ export class PGLiteEngine implements BrainEngine {
            p.slug, p.id as page_id, p.title, p.type, p.source_id,
            p.effective_date, p.effective_date_source,
            cc.id as chunk_id, cc.chunk_index, cc.chunk_text, cc.chunk_source,
-           ts_rank(cc.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score,
+           ts_rank(cc.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score,
            CASE WHEN p.updated_at < (
              SELECT MAX(te.created_at) FROM timeline_entries te WHERE te.page_id = p.id
            ) THEN true ELSE false END AS stale
          FROM content_chunks cc
          JOIN pages p ON p.id = cc.page_id
          JOIN sources s ON s.id = p.source_id
-         WHERE cc.search_vector @@ ${ftsQueryExpr} ${detailFilter}${extraFilter} ${hardExcludeClause} ${visibilityClause}
+         WHERE cc.search_vector @@ websearch_to_tsquery('${ftsLang}', $1) ${detailFilter}${extraFilter} ${hardExcludeClause} ${visibilityClause}
            -- v0.27.1: hide image rows from default text-keyword search so
            -- OCR text doesn't drown text-page hits. Image-similarity queries
            -- run a separate vector path on embedding_image.
@@ -1715,10 +1712,7 @@ export class PGLiteEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query, limit, offset];
     let extraFilter = '';
     if (opts?.type) {
@@ -1766,7 +1760,7 @@ export class PGLiteEngine implements BrainEngine {
          COALESCE(rep.chunk_index, 0) as chunk_index,
          COALESCE(rep.chunk_text, '') as chunk_text,
          COALESCE(rep.chunk_source, 'compiled_truth') as chunk_source,
-         ts_rank_cd(p.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score,
+         ts_rank_cd(p.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score,
          CASE WHEN p.updated_at < (
            SELECT MAX(te.created_at) FROM timeline_entries te WHERE te.page_id = p.id
          ) THEN true ELSE false END AS stale
@@ -1781,7 +1775,7 @@ export class PGLiteEngine implements BrainEngine {
          ORDER BY (cc.chunk_source = 'compiled_truth') DESC, cc.chunk_index ASC
          LIMIT 1
        ) rep ON true
-       WHERE p.search_vector @@ ${ftsQueryExpr}
+       WHERE p.search_vector @@ websearch_to_tsquery('${ftsLang}', $1)
          ${extraFilter} ${hardExcludeClause} ${visibilityClause}
        ORDER BY score DESC, p.id ASC
        LIMIT $2 OFFSET $3`;
@@ -1968,8 +1962,6 @@ export class PGLiteEngine implements BrainEngine {
       });
     }
 
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query, limit, offset];
     let extraFilter = '';
     if (opts?.language) {
@@ -2004,21 +1996,20 @@ export class PGLiteEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
     const { rows } = await this.db.query(
       `SELECT
          p.slug, p.id as page_id, p.title, p.type, p.source_id,
          p.effective_date, p.effective_date_source,
          cc.id as chunk_id, cc.chunk_index, cc.chunk_text, cc.chunk_source,
-         ts_rank(cc.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score,
+         ts_rank(cc.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score,
          CASE WHEN p.updated_at < (
            SELECT MAX(te.created_at) FROM timeline_entries te WHERE te.page_id = p.id
          ) THEN true ELSE false END AS stale
        FROM content_chunks cc
        JOIN pages p ON p.id = cc.page_id
        JOIN sources s ON s.id = p.source_id
-       WHERE cc.search_vector @@ ${ftsQueryExpr} ${detailFilter}${extraFilter} ${hardExcludeClause} ${visibilityClause}
+       WHERE cc.search_vector @@ websearch_to_tsquery('${ftsLang}', $1) ${detailFilter}${extraFilter} ${hardExcludeClause} ${visibilityClause}
        ORDER BY score DESC
        LIMIT $2 OFFSET $3`,
       params
@@ -4111,11 +4102,13 @@ export class PGLiteEngine implements BrainEngine {
   ): Promise<{ inserted: number; ids: number[] }> {
     if (rows.length === 0) return { inserted: 0, ids: [] };
 
-    // Single transaction so the v51 partial UNIQUE index can roll back the
-    // whole batch on constraint violation. Per-row INSERTs (not multi-row
-    // VALUES) keep the embedding-vs-no-embedding branching readable; batch
-    // sizes are small (5-30 rows per page in practice) so the loop overhead
-    // is negligible vs the embedding compute cost.
+    // Single transaction; per-row INSERTs (not multi-row VALUES) keep the
+    // embedding-vs-no-embedding branching readable; batch sizes are small
+    // (5-30 rows per page in practice) so the loop overhead is negligible
+    // vs the embedding compute cost. #2044: ON CONFLICT DO NOTHING on the
+    // v51 partial UNIQUE index makes a residual fence/DB row_num collision
+    // skip that row instead of rolling back the whole batch (parity with
+    // postgres-engine.ts).
     const ids = await this.db.transaction(async (tx) => {
       const out: number[] = [];
       for (const input of rows) {
@@ -4158,7 +4151,11 @@ export class PGLiteEngine implements BrainEngine {
                  $14, $15,
                  $16, $17, $18, $19,
                  $20
-               ) RETURNING id`
+               )
+               ON CONFLICT (source_id, source_markdown_slug, row_num)
+                 WHERE row_num IS NOT NULL
+                 DO NOTHING
+               RETURNING id`
             : `INSERT INTO facts (
                  source_id, entity_slug, fact, kind, visibility, notability, context,
                  valid_from, valid_until, source, source_session, confidence,
@@ -4172,12 +4169,16 @@ export class PGLiteEngine implements BrainEngine {
                  $15, $16,
                  $17, $18, $19, $20,
                  $21
-               ) RETURNING id`,
+               )
+               ON CONFLICT (source_id, source_markdown_slug, row_num)
+                 WHERE row_num IS NOT NULL
+                 DO NOTHING
+               RETURNING id`,
           embedStr === null
             ? [ctx.source_id, entitySlug, input.fact, kind, visibility, notability, context, validFrom, validUntil, input.source, sourceSession, confidence, embeddedAt, input.row_num, input.source_markdown_slug, claimMetric, claimValue, claimUnit, claimPeriod, eventType]
             : [ctx.source_id, entitySlug, input.fact, kind, visibility, notability, context, validFrom, validUntil, input.source, sourceSession, confidence, embedStr, embeddedAt, input.row_num, input.source_markdown_slug, claimMetric, claimValue, claimUnit, claimPeriod, eventType],
         );
-        out.push(ins.rows[0].id);
+        if (ins.rows[0]) out.push(ins.rows[0].id);
       }
       return out;
     });

src/core/postgres-engine.ts

@@ -64,7 +64,7 @@ import { ConnectionManager } from './connection-manager.ts';
 import { logConnectionEvent } from './connection-audit.ts';
 import { validateSlug, contentHash, rowToPage, rowToStalePage, rowToChunk, rowToSearchResult, parseEmbedding, tryParseEmbedding, takeRowToTake, takeHitRowToHit, isUndefinedTableError, warnOncePerProcess } from './utils.ts';
 import { resolveBoostMap, resolveHardExcludes } from './search/source-boost.ts';
-import { buildSourceFactorCase, buildHardExcludeClause, buildVisibilityClause, buildRecencyComponentSql, buildBestPerPagePoolCte, buildOrFallbackWebsearchQuery, buildWebsearchQueryExpr } from './search/sql-ranking.ts';
+import { buildSourceFactorCase, buildHardExcludeClause, buildVisibilityClause, buildRecencyComponentSql, buildBestPerPagePoolCte, buildOrFallbackWebsearchQuery } from './search/sql-ranking.ts';
 import { DEFAULT_EMBEDDING_MODEL, DEFAULT_EMBEDDING_DIMENSIONS } from './ai/defaults.ts';
 import { DELETE_BATCH_SIZE } from './engine-constants.ts';
 
@@ -1691,8 +1691,6 @@ export class PostgresEngine implements BrainEngine {
     const hardExcludePrefixes = resolveHardExcludes(opts?.exclude_slug_prefixes, opts?.include_slug_prefixes);
     const hardExcludeClause = buildHardExcludeClause('p.slug', hardExcludePrefixes);
 
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query];
     let typeClause = '';
     if (type) {
@@ -1763,7 +1761,6 @@ export class PostgresEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
     const rawQuery = `
       WITH ranked_chunks AS (
@@ -1771,11 +1768,11 @@ export class PostgresEngine implements BrainEngine {
           p.slug, p.id as page_id, p.title, p.type, p.source_id,
           p.effective_date, p.effective_date_source,
           cc.id as chunk_id, cc.chunk_index, cc.chunk_text, cc.chunk_source,
-          ts_rank(cc.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score
+          ts_rank(cc.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score
         FROM content_chunks cc
         JOIN pages p ON p.id = cc.page_id
         JOIN sources s ON s.id = p.source_id
-        WHERE cc.search_vector @@ ${ftsQueryExpr}
+        WHERE cc.search_vector @@ websearch_to_tsquery('${ftsLang}', $1)
           ${typeClause}
           ${typesClause}
           ${excludeSlugsClause}
@@ -1866,10 +1863,7 @@ export class PostgresEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query];
     let typeClause = '';
     if (opts?.type) {
@@ -1929,7 +1923,7 @@ export class PostgresEngine implements BrainEngine {
         COALESCE(rep.chunk_index, 0) as chunk_index,
         COALESCE(rep.chunk_text, '') as chunk_text,
         COALESCE(rep.chunk_source, 'compiled_truth') as chunk_source,
-        ts_rank_cd(p.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score,
+        ts_rank_cd(p.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score,
         false AS stale
       FROM pages p
       JOIN sources s ON s.id = p.source_id
@@ -1942,7 +1936,7 @@ export class PostgresEngine implements BrainEngine {
         ORDER BY (cc.chunk_source = 'compiled_truth') DESC, cc.chunk_index ASC
         LIMIT 1
       ) rep ON true
-      WHERE p.search_vector @@ ${ftsQueryExpr}
+      WHERE p.search_vector @@ websearch_to_tsquery('${ftsLang}', $1)
         ${typeClause}
         ${typesClause}
         ${excludeSlugsClause}
@@ -2006,8 +2000,6 @@ export class PostgresEngine implements BrainEngine {
     const hardExcludePrefixes = resolveHardExcludes(opts?.exclude_slug_prefixes, opts?.include_slug_prefixes);
     const hardExcludeClause = buildHardExcludeClause('p.slug', hardExcludePrefixes);
 
-    // #2380: slash-bearing queries match both the split-word and literal
-    // slash forms — see buildWebsearchQueryExpr in ./search/sql-ranking.ts.
     const params: unknown[] = [query];
     let typeClause = '';
     if (type) {
@@ -2068,19 +2060,18 @@ export class PostgresEngine implements BrainEngine {
     // FTS config name (e.g. 'english', 'pt_br'). Validated by getFtsLanguage()
     // — safe to interpolate into raw SQL.
     const ftsLang = getFtsLanguage();
-    const ftsQueryExpr = buildWebsearchQueryExpr(ftsLang, '$1', query);
 
     const rawQuery = `
       SELECT
         p.slug, p.id as page_id, p.title, p.type, p.source_id,
         p.effective_date, p.effective_date_source,
         cc.id as chunk_id, cc.chunk_index, cc.chunk_text, cc.chunk_source,
-        ts_rank(cc.search_vector, ${ftsQueryExpr}) * ${sourceFactorCase} AS score,
+        ts_rank(cc.search_vector, websearch_to_tsquery('${ftsLang}', $1)) * ${sourceFactorCase} AS score,
         false AS stale
       FROM content_chunks cc
       JOIN pages p ON p.id = cc.page_id
       JOIN sources s ON s.id = p.source_id
-      WHERE cc.search_vector @@ ${ftsQueryExpr}
+      WHERE cc.search_vector @@ websearch_to_tsquery('${ftsLang}', $1)
         ${typeClause}
         ${typesClause}
         ${excludeSlugsClause}
@@ -4311,10 +4302,12 @@ export class PostgresEngine implements BrainEngine {
     // ONCE per process so the cast matches the actual column type
     // (halfvec vs vector). The probe is cached after first call.
     const castSuffix = await this.resolveFactsEmbeddingCast();
-    // Single transaction so the v51 partial UNIQUE index can roll back
-    // the whole batch on constraint violation. Per-row INSERTs (not
-    // multi-row VALUES) keep the embedding-vs-no-embedding branching
-    // readable; batch sizes are small (5-30 rows per page in practice).
+    // Single transaction; per-row INSERTs (not multi-row VALUES) keep the
+    // embedding-vs-no-embedding branching readable; batch sizes are small
+    // (5-30 rows per page in practice). #2044: ON CONFLICT DO NOTHING on
+    // the v51 partial UNIQUE index makes a residual fence/DB row_num
+    // collision skip that row instead of rolling back the whole batch —
+    // the fence stays system-of-record and reconciliation catches up.
     // No supersede flow in this path — fence reconciliation is the
     // canonical source-of-truth direction, not the consolidator path.
     const ids = await sql.begin(async (tx) => {
@@ -4355,9 +4348,13 @@ export class PostgresEngine implements BrainEngine {
             ${input.row_num}, ${input.source_markdown_slug},
             ${claimMetric}, ${claimValue}, ${claimUnit}, ${claimPeriod},
             ${eventType}
-          ) RETURNING id
+          )
+          ON CONFLICT (source_id, source_markdown_slug, row_num)
+            WHERE row_num IS NOT NULL
+            DO NOTHING
+          RETURNING id
         `;
-        out.push(Number(ins[0].id));
+        if (ins[0]) out.push(Number(ins[0].id));
       }
       return out;
     });

src/core/search/sql-ranking.ts

@@ -251,28 +251,6 @@ export function buildOrFallbackWebsearchQuery(query: string): string | null {
   return tokens.join(' OR ');
 }
 
-/**
- * #2380: FTS query expression for slash-bearing queries. Postgres' default
- * text-search parser classifies `foo/bar` as a single `file`-alias lexeme —
- * on BOTH the query side and the index side. So a raw `foo/bar` query only
- * matched documents carrying the identical joined lexeme (literal paths),
- * and a slash-split query only matches documents whose text had the words
- * separated. Neither form alone covers both document shapes; OR the two
- * parses so a slash query matches prose ("foo and bar", stemmed, AND
- * semantics) AND literal slash forms ("src/core/x.ts") alike.
- *
- * Slash-free queries return the plain single-parse expression — byte-
- * identical SQL and identical ts_rank to the historical behavior.
- *
- * `ftsLang` is validated by getFtsLanguage() (safe to interpolate);
- * `param` is a `$N` placeholder, never user text.
- */
-export function buildWebsearchQueryExpr(ftsLang: string, param: string, query: string): string {
-  const plain = `websearch_to_tsquery('${ftsLang}', ${param})`;
-  if (!query.includes('/')) return plain;
-  return `(websearch_to_tsquery('${ftsLang}', translate(${param}, '/', ' ')) || ${plain})`;
-}
-
 // ============================================================
 // v0.29.1 — Recency component SQL builder
 // ============================================================

test/ai/no-batch-cap-suppression.serial.test.ts

@@ -28,8 +28,8 @@ describe('v0.32 #779: no_batch_cap suppresses the missing-max_batch_tokens warni
     resetGateway();
   });
 
-  test('LiteLLM and llama-server declare no_batch_cap: true', () => {
-    for (const id of ['litellm', 'llama-server']) {
+  test('Ollama, LiteLLM, llama-server all declare no_batch_cap: true', () => {
+    for (const id of ['ollama', 'litellm', 'llama-server']) {
       const r = getRecipe(id);
       expect(r, `${id} not registered`).toBeDefined();
       expect(
@@ -39,18 +39,6 @@ describe('v0.32 #779: no_batch_cap suppresses the missing-max_batch_tokens warni
     }
   });
 
-  test('#2552: Ollama declares a conservative static batch cap, not no_batch_cap', () => {
-    // A CPU-only Ollama box wedges when a whole page ships in one request;
-    // Ollama never returns a token-limit error so the recursive-halving
-    // safety net can't fire. The pre-split cap is the only guard.
-    const r = getRecipe('ollama');
-    expect(r).toBeDefined();
-    const e = r!.touchpoints.embedding!;
-    expect(e.no_batch_cap).toBeUndefined();
-    expect(e.max_batch_tokens).toBe(4096);
-    expect(e.chars_per_token).toBe(2);
-  });
-
   test('configureGateway does NOT warn for ollama/litellm/llama-server', () => {
     warnSpy.mockClear();
     resetGateway();

test/embed-local-endpoint.serial.test.ts

@@ -1,118 +0,0 @@
-/**
- * #2552: cloud-tuned embedding defaults silently wedge CPU-only local
- * endpoints (Ollama). Three-part fix under test:
- *
- *  1. `isLocalEmbeddingEndpoint()` — gateway helper detecting local
- *     inference servers (ollama / llama-server recipes, localhost base URL).
- *  2. `resolveEmbedConcurrency()` — embed auto-caps the 20-worker fan-out
- *     at LOCAL_EMBED_CONCURRENCY_CAP for local endpoints unless the
- *     operator set GBRAIN_EMBED_CONCURRENCY explicitly.
- *  3. `computeEmbedConcurrencyCheck()` — doctor warns when an explicit env
- *     override fans out against a local endpoint.
- *
- * Serial: mutates process.env and the module-global gateway config.
- */
-
-import { afterAll, afterEach, describe, expect, test } from 'bun:test';
-import {
-  configureGateway,
-  resetGateway,
-  isLocalEmbeddingEndpoint,
-  LOCAL_EMBED_CONCURRENCY_CAP,
-} from '../src/core/ai/gateway.ts';
-import { resolveEmbedConcurrency } from '../src/commands/embed.ts';
-import { computeEmbedConcurrencyCheck } from '../src/commands/doctor.ts';
-
-const SAVED_ENV = process.env.GBRAIN_EMBED_CONCURRENCY;
-
-afterEach(() => {
-  resetGateway();
-  if (SAVED_ENV === undefined) delete process.env.GBRAIN_EMBED_CONCURRENCY;
-  else process.env.GBRAIN_EMBED_CONCURRENCY = SAVED_ENV;
-});
-
-afterAll(() => {
-  resetGateway();
-});
-
-describe('#2552 isLocalEmbeddingEndpoint', () => {
-  test('false when the gateway is not configured (fail-open to cloud behavior)', () => {
-    resetGateway();
-    expect(isLocalEmbeddingEndpoint()).toBe(false);
-  });
-
-  test('true for the ollama recipe', () => {
-    configureGateway({ embedding_model: 'ollama:nomic-embed-text', env: {} });
-    expect(isLocalEmbeddingEndpoint()).toBe(true);
-  });
-
-  test('true for the llama-server recipe', () => {
-    configureGateway({ embedding_model: 'llama-server:my-gguf', env: {} });
-    expect(isLocalEmbeddingEndpoint()).toBe(true);
-  });
-
-  test('false for a cloud recipe', () => {
-    configureGateway({
-      embedding_model: 'openai:text-embedding-3-small',
-      env: { OPENAI_API_KEY: 'fake' },
-    });
-    expect(isLocalEmbeddingEndpoint()).toBe(false);
-  });
-
-  test('true when a cloud recipe base URL is explicitly pointed at localhost', () => {
-    configureGateway({
-      embedding_model: 'openai:text-embedding-3-small',
-      env: { OPENAI_API_KEY: 'fake' },
-      base_urls: { openai: 'http://localhost:8080/v1' },
-    });
-    expect(isLocalEmbeddingEndpoint()).toBe(true);
-  });
-});
-
-describe('#2552 resolveEmbedConcurrency', () => {
-  test('caps at LOCAL_EMBED_CONCURRENCY_CAP for a local endpoint when env is unset', () => {
-    delete process.env.GBRAIN_EMBED_CONCURRENCY;
-    configureGateway({ embedding_model: 'ollama:nomic-embed-text', env: {} });
-    expect(resolveEmbedConcurrency()).toBe(LOCAL_EMBED_CONCURRENCY_CAP);
-  });
-
-  test('explicit env override always wins, even against a local endpoint', () => {
-    process.env.GBRAIN_EMBED_CONCURRENCY = '10';
-    configureGateway({ embedding_model: 'ollama:nomic-embed-text', env: {} });
-    expect(resolveEmbedConcurrency()).toBe(10);
-  });
-
-  test('cloud endpoints keep the historical default of 20', () => {
-    delete process.env.GBRAIN_EMBED_CONCURRENCY;
-    configureGateway({ env: { OPENAI_API_KEY: 'fake' } });
-    expect(resolveEmbedConcurrency()).toBe(20);
-  });
-
-  test('pacing only ever lowers concurrency', () => {
-    delete process.env.GBRAIN_EMBED_CONCURRENCY;
-    configureGateway({ embedding_model: 'ollama:nomic-embed-text', env: {} });
-    expect(resolveEmbedConcurrency(1)).toBe(1);
-    expect(resolveEmbedConcurrency(16)).toBe(LOCAL_EMBED_CONCURRENCY_CAP);
-  });
-});
-
-describe('#2552 computeEmbedConcurrencyCheck (doctor)', () => {
-  test('ok for non-local endpoints', () => {
-    expect(computeEmbedConcurrencyCheck(false, '20', 2).status).toBe('ok');
-  });
-
-  test('warn when an explicit override exceeds the local cap', () => {
-    const check = computeEmbedConcurrencyCheck(true, '20', 2);
-    expect(check.status).toBe('warn');
-    expect(check.message).toContain('GBRAIN_EMBED_CONCURRENCY=20');
-  });
-
-  test('ok when env is unset against a local endpoint (auto-cap applies)', () => {
-    expect(computeEmbedConcurrencyCheck(true, undefined, 2).status).toBe('ok');
-  });
-
-  test('ok when the override is at or under the cap', () => {
-    expect(computeEmbedConcurrencyCheck(true, '2', 2).status).toBe('ok');
-    expect(computeEmbedConcurrencyCheck(true, '1', 2).status).toBe('ok');
-  });
-});

test/extract-facts-phase.test.ts

@@ -304,7 +304,9 @@ describe('runExtractFacts — empty-fence guard (Codex R2-#7)', () => {
     expect(r.legacyRowsPending).toBe(1);
     expect(r.factsInserted).toBe(0);
     expect(r.factsDeleted).toBe(0);
-    expect(r.warnings.some(w => w.includes('apply-migrations'))).toBe(true);
+    // #1867: the remedy hint points at the re-runnable backfill command,
+    // not the one-shot v0_32_2 migration (which the ledger never re-runs).
+    expect(r.warnings.some(w => w.includes('gbrain facts fence-backfill'))).toBe(true);
 
     // Legacy row was NOT touched.
     // eslint-disable-next-line @typescript-eslint/no-explicit-any

test/facts-fence-backfill-command.test.ts

@@ -0,0 +1,133 @@
+/**
+ * #1867 — `gbrain facts fence-backfill` command tests.
+ *
+ * The command re-runs the (idempotent) v0_32_2 phase B on demand so
+ * row_num-NULL backlogs — remote extract_facts deposits that predate
+ * the fence-write backstop — can be cleared without re-running the
+ * one-shot migration. Real PGLite + real tempdir filesystem.
+ */
+
+import { describe, test, expect, beforeAll, afterAll, beforeEach } from 'bun:test';
+import { mkdtempSync, rmSync, existsSync, readFileSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+
+import { PGLiteEngine } from '../src/core/pglite-engine.ts';
+import { runFactsCommand } from '../src/commands/facts.ts';
+import { phaseBFenceFacts } from '../src/commands/migrations/v0_32_2.ts';
+
+let engine: PGLiteEngine;
+let brainDir: string;
+
+beforeAll(async () => {
+  engine = new PGLiteEngine();
+  await engine.connect({});
+  await engine.initSchema();
+});
+
+afterAll(async () => {
+  await engine.disconnect();
+  try {
+    if (brainDir) rmSync(brainDir, { recursive: true, force: true });
+  } catch { /* best-effort */ }
+});
+
+beforeEach(async () => {
+  brainDir = mkdtempSync(join(tmpdir(), 'facts-backfill-cmd-test-'));
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  await (engine as any).db.query('DELETE FROM facts');
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  await (engine as any).db.query(
+    `UPDATE sources SET local_path = $1 WHERE id = 'default'`,
+    [brainDir],
+  );
+});
+
+async function seedLegacyFact(fact: string): Promise<void> {
+  // The row_num-NULL shape a remote extract_facts deposit leaves behind
+  // when it lands via the legacy DB-only path.
+  // eslint-disable-next-line @typescript-eslint/no-explicit-any
+  await (engine as any).db.query(
+    `INSERT INTO facts (source_id, entity_slug, fact, kind, visibility, notability,
+                        valid_from, source, confidence)
+     VALUES ('default', 'people/alice', $1, 'fact', 'private', 'medium',
+             now(), 'mcp:extract_facts', 1.0)`,
+    [fact],
+  );
+}
+
+describe('gbrain facts fence-backfill', () => {
+  test('fences row_num-NULL rows and stamps the DB', async () => {
+    await seedLegacyFact('Deposited remotely');
+
+    await runFactsCommand(engine, ['fence-backfill']);
+
+    // The fence exists on disk with the claim.
+    const filePath = join(brainDir, 'people/alice.md');
+    expect(existsSync(filePath)).toBe(true);
+    expect(readFileSync(filePath, 'utf-8')).toContain('Deposited remotely');
+
+    // The backlog is cleared: no row_num-NULL rows remain.
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    const rows = await (engine as any).db.query(
+      'SELECT row_num, source_markdown_slug FROM facts',
+    );
+    expect(rows.rows).toHaveLength(1);
+    expect(rows.rows[0].row_num).toBe(1);
+    expect(rows.rows[0].source_markdown_slug).toBe('people/alice');
+  });
+
+  test('re-run is a no-op (idempotent)', async () => {
+    await seedLegacyFact('Deposited remotely');
+    await runFactsCommand(engine, ['fence-backfill']);
+    await runFactsCommand(engine, ['fence-backfill']);
+
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    const rows = await (engine as any).db.query('SELECT id FROM facts');
+    expect(rows.rows).toHaveLength(1);
+    const body = readFileSync(join(brainDir, 'people/alice.md'), 'utf-8');
+    expect(body.match(/Deposited remotely/g)).toHaveLength(1);
+  });
+
+  test('--dry-run reports without writing', async () => {
+    await seedLegacyFact('Deposited remotely');
+    await runFactsCommand(engine, ['fence-backfill', '--dry-run']);
+
+    expect(existsSync(join(brainDir, 'people/alice.md'))).toBe(false);
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    const rows = await (engine as any).db.query(
+      'SELECT row_num FROM facts',
+    );
+    expect(rows.rows[0].row_num).toBeNull();
+  });
+
+  test('diverged page: appends past the DB row_num max instead of colliding (#2044 class)', async () => {
+    // The #2044 divergence shape the backfill must survive: the DB already
+    // holds stamped rows 1..3 for the slug (legacy wrong-path fence write),
+    // but the fence at the resolved path is missing. Fence-max+1 (= 1) would
+    // collide with the stamped rows on the post-rename UPDATE, failing the
+    // page and leaving the renamed fence disagreeing with the DB.
+    for (let n = 1; n <= 3; n++) {
+      // eslint-disable-next-line @typescript-eslint/no-explicit-any
+      await (engine as any).db.query(
+        `INSERT INTO facts (source_id, entity_slug, fact, kind, visibility, notability,
+                            valid_from, source, confidence, row_num, source_markdown_slug)
+         VALUES ('default', 'people/alice', $1, 'fact', 'private', 'medium',
+                 now(), 'mcp:extract_facts', 1.0, $2, 'people/alice')`,
+        [`stamped ${n}`, n],
+      );
+    }
+    await seedLegacyFact('Deposited remotely');
+
+    const result = await phaseBFenceFacts(engine, { dryRun: false });
+    expect(result.status).toBe('complete');
+
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    const rows = await (engine as any).db.query(
+      `SELECT row_num FROM facts WHERE fact = 'Deposited remotely'`,
+    );
+    expect(rows.rows[0].row_num).toBe(4);
+    const body = readFileSync(join(brainDir, 'people/alice.md'), 'utf-8');
+    expect(body).toContain('| 4 | Deposited remotely |');
+  });
+});

test/fence-write.test.ts

@@ -292,6 +292,49 @@ describe('lookupSourceLocalPath', () => {
   });
 });
 
+describe('writeFactsToFence — fence/DB divergence (#2044)', () => {
+  test('seeds row_num past the DB max when the fence lags the DB', async () => {
+    // Simulate the divergence class from #2044: DB rows were stamped with
+    // row_nums against a fence written at a path this code no longer reads
+    // (e.g. the pre-"Local patch 2026-06-11" wrong-path writes). The page
+    // on disk has NO fence, but the DB already holds row_num 1..3 for the
+    // slug. Pre-fix, the next deposit re-assigned row_num=1 from fence
+    // text alone and the whole insertFacts batch failed on
+    // idx_facts_fence_key.
+    for (let n = 1; n <= 3; n++) {
+      // eslint-disable-next-line @typescript-eslint/no-explicit-any
+      await (engine as any).db.query(
+        `INSERT INTO facts (source_id, entity_slug, fact, kind, visibility, notability,
+                            valid_from, source, confidence, row_num, source_markdown_slug)
+         VALUES ('default', 'people/dana', $1, 'fact', 'private', 'medium',
+                 now(), 'mcp:extract_facts', 1.0, $2, 'people/dana')`,
+        [`old claim ${n}`, n],
+      );
+    }
+
+    const result = await writeFactsToFence(
+      engine,
+      { sourceId: 'default', localPath: brainDir, slug: 'people/dana' },
+      [baseInput({ fact: 'second deposit' })],
+    );
+
+    expect(result.fenceWriteFailed).toBeUndefined();
+    expect(result.inserted).toBe(1);
+
+    // The new row landed PAST the DB max, not at fence-max+1 (= 1).
+    // eslint-disable-next-line @typescript-eslint/no-explicit-any
+    const rows = await (engine as any).db.query(
+      'SELECT row_num FROM facts WHERE id = $1',
+      [result.ids[0]],
+    );
+    expect(rows.rows[0].row_num).toBe(4);
+
+    // And the on-disk fence carries the same row_num — fence and DB agree.
+    const body = readFileSync(join(brainDir, 'people/dana.md'), 'utf-8');
+    expect(body).toContain('| 4 | second deposit |');
+  });
+});
+
 // Cleanup any leftover tempdirs after the whole suite.
 afterAll(() => {
   // No-op: each test cleaned up via the beforeEach; this is a safety net.

test/insert-facts-batch.test.ts

@@ -5,7 +5,7 @@
  *   - Batch insert N rows persists row_num + source_markdown_slug
  *   - Empty batch is a no-op
  *   - Returns ids in input-order
- *   - v51 partial UNIQUE index rolls back the whole batch on a collision
+ *   - v51 partial UNIQUE collision skips only the colliding row (#2044)
  *   - deleteFactsForPage scopes by (source_id, source_markdown_slug);
  *     never touches other pages or pre-v51 NULL-source_markdown_slug rows
  *   - deleteFactsForPage on an empty page returns deleted:0 (idempotent)
@@ -135,30 +135,29 @@ describe('engine.insertFacts — batch insert', () => {
     });
   });
 
-  test('v51 partial UNIQUE index rolls back the whole batch on collision', async () => {
+  test('v51 partial UNIQUE collision skips ONLY the colliding row (#2044)', async () => {
     // Seed row #1 first.
     await engine.insertFacts([fixtureFact(1, { fact: 'seeded' })], { source_id: 'default' });
 
-    // Now try to batch-insert rows that include a colliding row_num=1.
-    let threw = false;
-    try {
-      await engine.insertFacts(
-        [
-          fixtureFact(2, { fact: 'second' }),
-          fixtureFact(1, { fact: 'collides' }),  // row_num=1 on same (source_id, source_markdown_slug)
-          fixtureFact(3, { fact: 'third' }),
-        ],
-        { source_id: 'default' },
-      );
-    } catch {
-      threw = true;
-    }
-    expect(threw).toBe(true);
+    // Batch-insert rows that include a colliding row_num=1. Pre-#2044 this
+    // threw and rolled back the whole batch, making a second remote
+    // extract_facts deposit to an already-fenced page a hard failure. Now
+    // ON CONFLICT DO NOTHING skips the colliding row and keeps the rest.
+    const result = await engine.insertFacts(
+      [
+        fixtureFact(2, { fact: 'second' }),
+        fixtureFact(1, { fact: 'collides' }),  // row_num=1 on same (source_id, source_markdown_slug)
+        fixtureFact(3, { fact: 'third' }),
+      ],
+      { source_id: 'default' },
+    );
+    expect(result.inserted).toBe(2);
+    expect(result.ids).toHaveLength(2);
 
-    // Verify the transaction rolled back — only the seeded row should remain.
+    // The seeded row survives untouched; the colliding claim is skipped.
     // eslint-disable-next-line @typescript-eslint/no-explicit-any
     const rows = await (engine as any).db.query('SELECT fact FROM facts ORDER BY id');
-    expect(rows.rows.map((r: { fact: string }) => r.fact)).toEqual(['seeded']);
+    expect(rows.rows.map((r: { fact: string }) => r.fact)).toEqual(['seeded', 'second', 'third']);
   });
 
   test('different source_markdown_slug values DO NOT collide on the same row_num', async () => {

test/pglite-engine.test.ts

@@ -216,67 +216,6 @@ describe('PGLiteEngine: Search', () => {
     expect(results.length).toBe(0);
   });
 
-  // Regression (#2380): queries containing `/` used to bypass FTS AND
-  // semantics. Postgres' default text-search parser classifies `foo/bar` as
-  // a `file`-alias token mapped to the `simple` dictionary, so it became a
-  // single un-stemmed lexeme `'foo/bar'` that never matches indexed text —
-  // the primary FTS pass returned 0 and the OR fallback took over, matching
-  // pages that contain EITHER term. searchKeyword/searchTitles now normalize
-  // `/` to whitespace before websearch_to_tsquery parses, so the primary
-  // AND pass matches directly.
-  test('searchKeyword: slash query matches with AND semantics, not OR fallback', async () => {
-    // Decoy shares only ONE of the two query terms ('enterprise').
-    await engine.putPage('concepts/enterprise-pricing', {
-      type: 'concept', title: 'Widget Pricing',
-      compiled_truth: 'Enterprise pricing for widgets.',
-    });
-    await engine.upsertChunks('concepts/enterprise-pricing', [
-      { chunk_index: 0, chunk_text: 'Enterprise pricing for widgets', chunk_source: 'compiled_truth' },
-    ]);
-
-    // Both terms co-occur only in the novamind chunk. Pre-fix this returned
-    // BOTH pages (primary pass zero-hit → OR fallback); post-fix the primary
-    // AND pass returns exactly the co-occurrence page.
-    const results = await engine.searchKeyword('NovaMind/enterprise');
-    expect(results.length).toBe(1);
-    expect(results[0].slug).toBe('companies/novamind');
-  });
-
-  test('searchTitles: slash query matches with AND semantics, not OR fallback', async () => {
-    await engine.putPage('companies/novamind-enterprise', {
-      type: 'company', title: 'NovaMind Enterprise Platform',
-      compiled_truth: 'Placeholder body.',
-    });
-    await engine.putPage('guides/enterprise-sales', {
-      type: 'concept', title: 'Enterprise Sales Guide',
-      compiled_truth: 'Placeholder body.',
-    });
-
-    // Pre-fix: `NovaMind/Enterprise` parsed as one file-alias lexeme → the
-    // primary title pass returned 0 and the OR fallback matched BOTH titles.
-    const results = await engine.searchTitles('NovaMind/Enterprise');
-    expect(results.length).toBe(1);
-    expect(results[0].slug).toBe('companies/novamind-enterprise');
-  });
-
-  test('searchKeyword: slash query still matches the literal slash form (file paths)', async () => {
-    // The INDEX side also emits the joined file-alias lexeme for literal
-    // `foo/bar` text, so a query normalized to split words alone would go
-    // blind to documents containing the literal slash form (paths, URLs).
-    // buildWebsearchQueryExpr ORs both parses; this pins the raw arm.
-    await engine.putPage('runbooks/widget-deploy', {
-      type: 'concept', title: 'Widget Deploy Runbook',
-      compiled_truth: 'Runbook for the acme/widget deployment pipeline.',
-    });
-    await engine.upsertChunks('runbooks/widget-deploy', [
-      { chunk_index: 0, chunk_text: 'Runbook for the acme/widget deployment pipeline', chunk_source: 'compiled_truth' },
-    ]);
-
-    const results = await engine.searchKeyword('acme/widget');
-    expect(results.length).toBe(1);
-    expect(results[0].slug).toBe('runbooks/widget-deploy');
-  });
-
   test('tsvector trigger populates search_vector on insert', async () => {
     // Verify the PL/pgSQL trigger fires and content_chunks.search_vector is
     // populated from chunk_text. v0.20.0 Cathedral II Layer 3 moved FTS from