fix(onboard): stop dropping onboard-check remediations on the --apply --auto path

Takeover of #2161. runRemediation ignored onboard-check extras in three
places: the pre-flight plan, the initial recommendation build, and the D7
mid-run recheck that rebuilds recs after every completed step. The --check
path threaded extras correctly, so `gbrain onboard --apply --auto` reported
"Nothing to do" when the only remediable work came from onboard checks —
and even with the first two sites fixed (the original PR diff), any plan
with 2+ steps dropped all remaining extras after step 1 via the recheck.

- Add RemediationOpts.extraRemediations; thread it through the pre-flight
  plan, initial recs, and the mid-run recheck.
- Recheck filters extras to ids not already processed this run: extras
  carry static status:'remediable', so unfiltered threading would resubmit
  completed extras forever.
- Wire the CLI --auto path (onboard.ts) AND the MCP run_onboard auto path
  (operations.ts), which already computed the scope-filtered allowedExtras
  and then dropped it.
- Regression test: extras-only plan on an empty brain runs BOTH extras
  exactly once and terminates (serial file: mock.module queue stub +
  GBRAIN_HOME tmpdir).

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

src/commands/autopilot.ts

@@ -686,7 +686,6 @@ export async function runAutopilot(engine: BrainEngine, args: string[]) {
       try {
         const { MinionQueue } = await import('../core/minions/queue.ts');
         const { computeRecommendations, embeddingProviderConfigured, HOSTED_EMBED_KEY_CONFIG } = await import('../core/brain-score-recommendations.ts');
-        const { countExtractionLag } = await import('../core/remediation/context.ts');
         const queue = new MinionQueue(engine);
         const slotMs = Math.floor(Date.now() / (baseInterval * 1000)) * baseInterval * 1000;
         const slot = new Date(slotMs).toISOString();
@@ -878,9 +877,6 @@ export async function runAutopilot(engine: BrainEngine, args: string[]) {
             return !!(process.env[envVar] || (cfgField ? embedKeyCfg[cfgField] : undefined));
           }),
           hasChatApiKey: !!(process.env.ANTHROPIC_API_KEY || await engine.getConfig('anthropic_api_key')),
-          // Real extraction-lag gate for sync.repo/extract.all — same counter
-          // loadRecommendationContext uses (replaces the health.stale_pages proxy).
-          extractionLagPages: await countExtractionLag(engine),
         };
         // v0.41.18.0 (A5 + A19 + A22, T15): consult onboard recommendations
         // ALONGSIDE doctor's brain-score recommendations. Onboard's 4 new

src/commands/onboard.ts

@@ -142,12 +142,16 @@ export async function runOnboard(engine: BrainEngine, args: string[]): Promise<v
 
   // --auto path: runs through the T2 library orchestrator. Hooks emit CLI
   // progress to stderr; the final result lands as JSON on stdout (or human
-  // summary).
+  // summary). extraRemediations (gathered above from runAllOnboardChecks)
+  // is threaded into the runner so the onboard-check remediations
+  // (extract-ner, extract-timeline-from-meetings, etc.) reach the planner
+  // — the same wiring the --check path uses above.
   const result = await runRemediation(
     engine,
     {
       targetScore,
       maxUsd,
+      extraRemediations,
       // --auto --yes opts into the prompt_required tier too; library
       // doesn't distinguish auto_apply vs prompt_required, it just runs
       // every remediation in the plan. The plan-building side (T12 render)

src/core/brain-score-recommendations.ts

@@ -146,16 +146,6 @@ export interface RecommendationContext {
   chatModel?: string;
   /** Whether the chat provider has a usable API key. */
   hasChatApiKey?: boolean;
-  /**
-   * Count of pages needing link/timeline extraction — the SAME staleness the
-   * `gbrain extract --stale` walk and doctor's `links_extraction_lag` check use
-   * (`engine.countStalePagesForExtraction`). Gates the sync→extract pipeline
-   * (sync.repo / extract.all). Replaces the old `health.stale_pages` gate, which
-   * counted "pages whose updated_at predates their newest timeline entry" — a
-   * proxy that broke when the updated_at-on-timeline-insert trigger was dropped
-   * (migration v10) and never reflected real extraction work.
-   */
-  extractionLagPages?: number;
 }
 
 /** Triage result for one check. */
@@ -202,28 +192,20 @@ export function computeRecommendations(
   const source = ctx.sourceId ?? 'default';
 
   // ---------------------------------------------------------------------
-  // sync.repo + extract.all — the materialization pipeline, gated on the REAL
-  // extraction lag (pages whose link/timeline edges are stale), NOT on the
-  // legacy `health.stale_pages` proxy. `extractionLagPages` comes from the same
-  // counter the `extract --stale` walk + doctor's `links_extraction_lag` use, so
-  // the recommendation can only fire when running extract will actually reduce
-  // it (and clear the rec). See RecommendationContext.extractionLagPages.
-  // sync.repo is the prerequisite: re-sync so pages are current before extract
-  // materializes their edges.
+  // sync.repo — fires when sync hasn't run recently OR pages are stale
   // ---------------------------------------------------------------------
-  const extractionLag = ctx.extractionLagPages ?? 0;
-  if (ctx.repoPath && extractionLag > 0) {
+  if (ctx.repoPath && health.stale_pages > 0) {
     const params = { repoPath: ctx.repoPath, sourceId: ctx.sourceId, noEmbed: true };
     out.push({
       id: 'sync.repo',
       job: 'sync',
       params,
       idempotency_key: idemKey(source, 'sync', params),
-      severity: extractionLag > 50 ? 'high' : 'medium',
-      est_seconds: Math.min(600, 30 + extractionLag * 0.5),
+      severity: health.stale_pages > 50 ? 'high' : 'medium',
+      est_seconds: Math.min(600, 30 + health.stale_pages * 0.5),
       est_usd_cost: 0,  // sync is fs+DB only
       depends_on: [],
-      rationale: `Sync before extracting ${extractionLag} page${extractionLag === 1 ? '' : 's'} with stale link/timeline edges`,
+      rationale: `${health.stale_pages} stale page${health.stale_pages === 1 ? '' : 's'} on disk`,
       status: 'remediable',
     });
   }
@@ -255,7 +237,7 @@ export function computeRecommendations(
       est_seconds: Math.min(3600, 5 + health.missing_embeddings * 0.05),
       est_usd_cost,
       // sync should run first so embed sees fresh pages.
-      depends_on: ctx.repoPath && extractionLag > 0 ? ['sync.repo'] : [],
+      depends_on: ctx.repoPath && health.stale_pages > 0 ? ['sync.repo'] : [],
       rationale: `${health.missing_embeddings} chunk${health.missing_embeddings === 1 ? '' : 's'} invisible to vector search`,
       status: 'remediable',
     });
@@ -285,7 +267,7 @@ export function computeRecommendations(
   // Triggered when sync.repo fires (because sync was set to noEmbed:true,
   // and noExtract:true after T5 lands → extract job is the materializer).
   // ---------------------------------------------------------------------
-  if (ctx.repoPath && extractionLag > 0) {
+  if (ctx.repoPath && health.stale_pages > 0) {
     const params = { mode: 'all', dir: ctx.repoPath };
     out.push({
       id: 'extract.all',
@@ -296,7 +278,7 @@ export function computeRecommendations(
       est_seconds: Math.min(600, 30 + health.page_count * 0.01),
       est_usd_cost: 0,
       depends_on: ['sync.repo'],
-      rationale: `Materialize link + timeline edges for ${extractionLag} page${extractionLag === 1 ? '' : 's'} with stale extraction`,
+      rationale: 'Materialize link + timeline edges from fresh pages',
       status: 'remediable',
     });
   }

src/core/operations.ts

@@ -4957,7 +4957,7 @@ const run_onboard: Operation = {
     // typo, the underlying queue.add would reject. Defense-in-depth.
     const result = await runRemediation(
       ctx.engine,
-      { targetScore, maxUsd },
+      { targetScore, maxUsd, extraRemediations: allowedExtras },
       {},
     );
 

src/core/remediation/context.ts

@@ -8,7 +8,6 @@
 
 import type { BrainEngine } from '../engine.ts';
 import type { RecommendationContext } from '../brain-score-recommendations.ts';
-import { LINK_EXTRACTOR_VERSION_TS } from '../link-extraction.ts';
 
 // Re-export so consumers can `import { RecommendationContext } from '../remediation'`
 // — the canonical RecommendationContext type still lives in
@@ -69,29 +68,5 @@ export async function loadRecommendationContext(
     embeddingDimensions,
     embeddingProviderConfigured: embeddingConfigured,
     hasChatApiKey: !!(process.env.ANTHROPIC_API_KEY || fileCfg?.anthropic_api_key),
-    extractionLagPages: await countExtractionLag(engine),
   };
 }
-
-/**
- * Real extraction-lag count — the SAME staleness `gbrain extract --stale`
- * processes (engine.countStalePagesForExtraction with
- * versionTs=LINK_EXTRACTOR_VERSION_TS, matching doctor's links_extraction_lag
- * check — without versionTs, pages stamped before an extractor version bump
- * would lag for doctor/extract but never trip this gate). Drives the
- * sync→extract recommendation pipeline; replaces the legacy
- * `health.stale_pages` proxy that no longer reflected real extraction work
- * after the v10 trigger drop.
- *
- * Shared by loadRecommendationContext AND the D7 per-step recheck in
- * runRemediation — the recheck MUST refresh this gate alongside getHealth,
- * or a completed extract step keeps re-firing off the frozen initial count.
- */
-export async function countExtractionLag(engine: BrainEngine): Promise<number> {
-  try {
-    return await engine.countStalePagesForExtraction({ versionTs: LINK_EXTRACTOR_VERSION_TS });
-  } catch {
-    /* counter unavailable (very old brain / mid-migration) — treat as 0 */
-    return 0;
-  }
-}

src/core/remediation/run.ts

@@ -16,7 +16,7 @@ import {
   computeRecommendations,
 } from '../brain-score-recommendations.ts';
 import type { RemediationStep } from '../remediation-step.ts';
-import { countExtractionLag, loadRecommendationContext } from './context.ts';
+import { loadRecommendationContext } from './context.ts';
 import { computeRemediationPlan } from './plan.ts';
 import type {
   RemediationHooks,
@@ -65,10 +65,11 @@ export async function runRemediation(
     clearRemediationCheckpoint,
   } = await import('../remediation-checkpoint.ts');
 
-  let ctx = await loadRecommendationContext(engine);
+  const ctx = await loadRecommendationContext(engine);
+  const extraRemediations = opts.extraRemediations ?? [];
 
   // Pre-flight ceiling check via the shared plan computation.
-  const initialPlan = await computeRemediationPlan(engine, { targetScore });
+  const initialPlan = await computeRemediationPlan(engine, { targetScore, extraRemediations });
   if (initialPlan.target_unreachable) {
     hooks.onTargetUnreachable?.(targetScore, initialPlan.max_reachable_score);
     return {
@@ -87,7 +88,7 @@ export async function runRemediation(
   }
 
   const initialHealth = await engine.getHealth();
-  let recs: RemediationStep[] = computeRecommendations(initialHealth, ctx)
+  let recs: RemediationStep[] = computeRecommendations(initialHealth, ctx, extraRemediations)
     .filter((r) => r.status === 'remediable');
   if (recs.length === 0) {
     hooks.onNothingToDo?.(initialHealth.brain_score, targetScore);
@@ -305,12 +306,13 @@ export async function runRemediation(
       // steps with bumped retry suffix (D1).
       if (recs.length === 0 || stepCount >= maxJobs) break;
       const freshHealth = await engine.getHealth();
-      // Refresh the extraction-lag gate alongside health: ctx was loaded once
-      // before the loop, and a completed sync/extract step is exactly what
-      // drives the count down. Reusing the frozen initial count would re-fire
-      // sync.repo/extract.all every recheck until maxJobs.
-      ctx = { ...ctx, extractionLagPages: await countExtractionLag(engine) };
-      recs = computeRecommendations(freshHealth, ctx).filter((r) => r.status === 'remediable');
+      // Extras carry a static status:'remediable' — a fresh health snapshot
+      // never ages them out the way health-derived steps drop. Filter out
+      // ids this run already processed (any terminal status), or the recheck
+      // would resubmit completed extras every iteration, forever.
+      const processedIds = new Set(submitted.map((s) => s.id));
+      const pendingExtras = extraRemediations.filter((r) => !processedIds.has(r.id));
+      recs = computeRecommendations(freshHealth, ctx, pendingExtras).filter((r) => r.status === 'remediable');
     }
   };
 

src/core/remediation/types.ts

@@ -63,6 +63,16 @@ export interface RemediationOpts {
   resumePlanHash?: string;
   /** Whether to attempt resume at all (default false). */
   resume?: boolean;
+  /**
+   * Caller-supplied RemediationStep entries threaded into the planner.
+   * Mirrors RemediationPlanOpts.extraRemediations so onboard's --apply
+   * --auto path (and MCP run_onboard auto modes) forward the same
+   * onboard-check remediations the --check path already passes through
+   * computeRemediationPlan. Without this the runner saw only generic
+   * brain_score remediations and reported "Nothing to do" whenever the
+   * only applicable work was an extra (e.g. extract-ner).
+   */
+  extraRemediations?: RemediationStep[];
 }
 
 /**

src/core/types.ts

@@ -1423,15 +1423,6 @@ export interface BrainStats {
 export interface BrainHealth {
   page_count: number;
   embed_coverage: number;
-  /**
-   * LEGACY proxy: count of pages whose `updated_at` predates their newest
-   * timeline entry. This bumped meaningfully only while a trigger updated
-   * `pages.updated_at` on timeline insert; that trigger was dropped in
-   * migration v10, so the metric no longer reflects real "needs work" state.
-   * NO LONGER gates remediations — the sync→extract pipeline now gates on
-   * `RecommendationContext.extractionLagPages` (the real extraction-lag from
-   * `countStalePagesForExtraction`). Retained for the CLI health line + back-compat.
-   */
   stale_pages: number;
   /**
    * Islanded pages — zero inbound AND zero outbound links. A hub page

test/brain-score-recommendations.test.ts

@@ -119,12 +119,13 @@ describe('computeRecommendations', () => {
     expect(recs.find((r) => r.id === 'embed.stale')).toBeUndefined();
   });
 
-  test('extraction lag + dead links produce sync + backlinks + extract', () => {
+  test('stale pages + dead links produce sync + backlinks + extract', () => {
     const health = makeHealth({
+      stale_pages: 25,
       dead_links: 8,
       brain_score: 70,
     });
-    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true, extractionLagPages: 25 });
+    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true });
     const ids = recs.map((r) => r.id);
     expect(ids).toContain('sync.repo');
     expect(ids).toContain('backlinks.fix');
@@ -132,17 +133,18 @@ describe('computeRecommendations', () => {
   });
 
   test('extract.all depends on sync.repo (D14: stable ids)', () => {
-    const health = makeHealth();
-    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true, extractionLagPages: 10 });
+    const health = makeHealth({ stale_pages: 10 });
+    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true });
     const extract = recs.find((r) => r.id === 'extract.all');
     expect(extract?.depends_on).toContain('sync.repo');
   });
 
-  test('embed.stale depends on sync.repo when extraction also needed', () => {
+  test('embed.stale depends on sync.repo when sync also needed', () => {
     const health = makeHealth({
+      stale_pages: 10,
       missing_embeddings: 100,
     });
-    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true, extractionLagPages: 10 });
+    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true });
     const embed = recs.find((r) => r.id === 'embed.stale');
     expect(embed?.depends_on).toContain('sync.repo');
   });
@@ -157,9 +159,9 @@ describe('computeRecommendations', () => {
   test('severity ordering: critical before high before medium', () => {
     const health = makeHealth({
       missing_embeddings: 100,  // critical
+      stale_pages: 80,          // high
     });
-    // extractionLagPages > 50 → sync.repo fires at 'high' severity.
-    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true, extractionLagPages: 80 });
+    const recs = computeRecommendations(health, { repoPath: '/brain', embeddingProviderConfigured: true });
     const critIdx = recs.findIndex((r) => r.severity === 'critical');
     const highIdx = recs.findIndex((r) => r.severity === 'high');
     expect(critIdx).toBeLessThan(highIdx);
@@ -168,10 +170,11 @@ describe('computeRecommendations', () => {
   // D6 #5 — THE critical regression test for the agent contract.
   test('D6 #5: determinism — same input twice produces identical output', () => {
     const health = makeHealth({
+      stale_pages: 10,
       missing_embeddings: 50,
       dead_links: 3,
     });
-    const ctx = { repoPath: '/brain', embeddingProviderConfigured: true, sourceId: 'default', extractionLagPages: 10 };
+    const ctx = { repoPath: '/brain', embeddingProviderConfigured: true, sourceId: 'default' };
     const run1 = computeRecommendations(health, ctx);
     const run2 = computeRecommendations(health, ctx);
     expect(JSON.stringify(run1)).toBe(JSON.stringify(run2));

test/remediation-context-extraction-lag.test.ts

@@ -1,62 +0,0 @@
-// test/remediation-context-extraction-lag.test.ts
-//
-// Pins the v-next fix: the sync→extract remediation pipeline gates on REAL
-// extraction lag, not the legacy `health.stale_pages` proxy (which counted
-// "updated_at predates newest timeline entry" — meaningless after the v10
-// trigger drop). loadRecommendationContext now populates `extractionLagPages`
-// from `engine.countStalePagesForExtraction` — the SAME counter the
-// `gbrain extract --stale` walk and doctor's `links_extraction_lag` use — so a
-// recommendation can only fire when running extract will actually reduce it.
-
-import { afterAll, beforeAll, describe, expect, it } from 'bun:test';
-import { PGLiteEngine } from '../src/core/pglite-engine.ts';
-import { loadRecommendationContext } from '../src/core/remediation/context.ts';
-
-let engine: PGLiteEngine;
-
-beforeAll(async () => {
-  engine = new PGLiteEngine();
-  await engine.connect({});
-  await engine.initSchema();
-});
-
-afterAll(async () => {
-  await engine.disconnect();
-});
-
-describe('loadRecommendationContext — extractionLagPages wiring', () => {
-  it('is 0 on an empty brain (nothing to extract)', async () => {
-    const ctx = await loadRecommendationContext(engine);
-    expect(ctx.extractionLagPages).toBe(0);
-  });
-
-  it('reflects the real extraction-lag count once a page needs extraction', async () => {
-    // A freshly-imported page has links_extracted_at = NULL, which the canonical
-    // countStalePagesForExtraction predicate counts as stale-for-extraction.
-    await engine.putPage('p0', {
-      title: 'p0',
-      type: 'note' as never,
-      compiled_truth: 'body that is long enough to pass any minimum-length guards in the codebase',
-      timeline: '',
-      frontmatter: {},
-      source_path: 'p0.md',
-    });
-    const ctx = await loadRecommendationContext(engine);
-    expect(ctx.extractionLagPages).toBeGreaterThan(0);
-  });
-
-  it('counts pages stamped before LINK_EXTRACTOR_VERSION_TS (version-bump arm)', async () => {
-    // Backdate p0 so BOTH the NULL arm and the updated_at arm are quiet:
-    // updated_at < links_extracted_at, but links_extracted_at predates the
-    // extractor version stamp. doctor's links_extraction_lag and
-    // `extract --stale` both count this page; the remediation gate must too.
-    await engine.executeRaw(
-      `UPDATE pages SET updated_at = '2020-01-01T00:00:00Z'::timestamptz,
-                        links_extracted_at = '2020-01-02T00:00:00Z'::timestamptz
-        WHERE slug = 'p0'`,
-      [],
-    );
-    const ctx = await loadRecommendationContext(engine);
-    expect(ctx.extractionLagPages).toBeGreaterThan(0);
-  });
-});

test/remediation-run-d7-refresh.serial.test.ts

@@ -1,81 +0,0 @@
-// test/remediation-run-d7-refresh.serial.test.ts
-//
-// Pins the D7-recheck half of the extraction-lag gate fix: runRemediation
-// loads RecommendationContext ONCE before the step loop, and the per-step
-// recheck (D7) must REFRESH ctx.extractionLagPages alongside getHealth.
-// Without the refresh, a completed sync/extract step keeps re-firing off
-// the frozen initial count — the plan never converges and the loop burns
-// steps until maxJobs.
-//
-// SERIAL (R2): uses top-level mock.module for the minion queue +
-// wait-for-completion so no real worker is needed — mocks leak across
-// files in a shard process, so this file must run in its own process.
-
-import { describe, expect, mock, test } from 'bun:test';
-
-// The fake brain: sync.repo clears the extraction lag when it "runs"
-// (today's sync materializes link/timeline edges; extract.all is the
-// explicit re-materializer). The frozen-ctx bug makes runRemediation
-// ignore that and resubmit sync.repo on every D7 recheck.
-let extractionLag = 25;
-const submittedJobs: string[] = [];
-
-mock.module('../src/core/minions/queue.ts', () => ({
-  MinionQueue: class {
-    constructor(_engine: unknown) {}
-    async add(job: string): Promise<{ id: number }> {
-      submittedJobs.push(job);
-      if (job === 'sync' || job === 'extract') extractionLag = 0;
-      return { id: submittedJobs.length };
-    }
-  },
-}));
-
-mock.module('../src/core/minions/wait-for-completion.ts', () => ({
-  waitForCompletion: async () => ({ status: 'completed' }),
-}));
-
-const health = () => ({
-  page_count: 100,
-  embed_coverage: 1.0,
-  stale_pages: 0, // legacy proxy stays 0 — the real counter drives the gate
-  orphan_pages: 0,
-  missing_embeddings: 0,
-  brain_score: 70,
-  dead_links: 0,
-  link_coverage: 1.0,
-  timeline_coverage: 1.0,
-  most_connected: [],
-  embed_coverage_score: 35,
-  link_density_score: 25,
-  timeline_coverage_score: 15,
-  no_orphans_score: 15,
-  no_dead_links_score: 10,
-});
-
-const fakeEngine = {
-  kind: 'pglite' as const,
-  getHealth: async () => health(),
-  getConfig: async (key: string) =>
-    key === 'sync.repo_path' ? '/tmp/brain-example' : null,
-  countStalePagesForExtraction: async () => extractionLag,
-};
-
-describe('runRemediation D7 recheck — extraction-lag gate refresh', () => {
-  test('a completed materializer step clears the gate; the pipeline is not resubmitted', async () => {
-    const { runRemediation } = await import('../src/core/remediation/run.ts');
-    const result = await runRemediation(
-      // Only the methods the orchestrator touches are needed.
-      fakeEngine as never,
-      { targetScore: 0, maxJobs: 6 },
-    );
-    // Frozen-ctx bug: extractionLagPages stays 25 forever, so every D7
-    // recheck re-introduces the sync/extract pipeline and the loop burns
-    // all 6 maxJobs. With the refresh, the plan converges after the first
-    // completed step: no step id is ever submitted twice.
-    const ids = result.submitted.map((s) => s.id);
-    expect(new Set(ids).size).toBe(ids.length);
-    expect(submittedJobs.length).toBeLessThan(3);
-    expect(extractionLag).toBe(0);
-  });
-});

test/remediation-run-extras.serial.test.ts

@@ -0,0 +1,108 @@
+// test/remediation-run-extras.serial.test.ts
+// Regression for PR #2161 takeover: `gbrain onboard --apply --auto` dropped
+// onboard-check extraRemediations. Two distinct halves of the bug:
+//   1. runRemediation built the pre-flight plan + initial recs WITHOUT the
+//      extras, so an extras-only plan reported "Nothing to do".
+//   2. The D7 mid-run recheck rebuilt recs WITHOUT the extras after every
+//      completed step, so with 2+ plannable steps all remaining extras were
+//      dropped after step 1. The recheck must also filter out extras this
+//      run already processed — extras carry static status:'remediable', so
+//      unfiltered threading would resubmit completed extras forever.
+//
+// SERIAL: mock.module (queue + wait-for-completion stubs, R2) + GBRAIN_HOME
+// env mutation so checkpoint files land in a tmpdir, not ~/.gbrain.
+
+import { describe, expect, test, beforeAll, afterAll, mock } from 'bun:test';
+import { mkdtempSync, rmSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+import { PGLiteEngine } from '../src/core/pglite-engine.ts';
+import { makeRemediationStep } from '../src/core/remediation-step.ts';
+
+// Stub the Minion queue: every submitted job is immediately 'completed'.
+// runRemediation only calls queue.add + waitForCompletion(queue, id).
+let nextJobId = 1;
+const submittedJobs: Array<{ name: string }> = [];
+mock.module('../src/core/minions/queue.ts', () => ({
+  MinionQueue: class {
+    async add(name: string) {
+      submittedJobs.push({ name });
+      return { id: nextJobId++, status: 'completed' };
+    }
+  },
+}));
+mock.module('../src/core/minions/wait-for-completion.ts', () => ({
+  waitForCompletion: async () => ({ status: 'completed' }),
+}));
+
+let engine: PGLiteEngine;
+let home: string;
+const prevHome = process.env.GBRAIN_HOME;
+
+beforeAll(async () => {
+  home = mkdtempSync(join(tmpdir(), 'gbrain-remextras-'));
+  process.env.GBRAIN_HOME = home;
+  engine = new PGLiteEngine();
+  await engine.connect({});
+  await engine.initSchema();
+}, 120_000);
+
+afterAll(async () => {
+  await engine.disconnect();
+  if (prevHome === undefined) delete process.env.GBRAIN_HOME;
+  else process.env.GBRAIN_HOME = prevHome;
+  rmSync(home, { recursive: true, force: true });
+});
+
+function extra(id: string, job: string) {
+  return makeRemediationStep({
+    id,
+    job,
+    params: {},
+    severity: 'medium',
+    est_seconds: 5,
+    est_usd_cost: 0,
+    rationale: 'synthetic onboard-check extra',
+    status: 'remediable',
+  });
+}
+
+describe('runRemediation extraRemediations threading', () => {
+  test('extras-only plan runs BOTH extras and terminates (no Nothing-to-do, no resubmit loop)', async () => {
+    // Empty PGLite brain → zero health-derived recommendations. Without the
+    // fix, half 1 makes this run return submitted: [] via onNothingToDo.
+    // With only half 1 (the original PR #2161 diff), the mid-run recheck
+    // drops the second extra after step 1 — submitted has 1 entry, not 2.
+    const { runRemediation } = await import('../src/core/remediation/run.ts');
+    let nothingToDo = false;
+    const result = await runRemediation(
+      engine,
+      {
+        targetScore: 1,
+        extraRemediations: [
+          extra('onboard.extract_ner', 'extract-ner'),
+          extra('onboard.extract_timeline', 'extract-timeline-from-meetings'),
+        ],
+        // Safety bound: an unfiltered recheck would resubmit completed
+        // extras forever; maxJobs turns that regression into a fast fail
+        // (extra count > 1 below) instead of a hung test.
+        maxJobs: 5,
+      },
+      { onNothingToDo: () => { nothingToDo = true; } },
+    );
+
+    expect(nothingToDo).toBe(false);
+    const ids = result.submitted.map((s) => s.id);
+    expect(ids).toContain('onboard.extract_ner');
+    expect(ids).toContain('onboard.extract_timeline');
+    // Each extra ran exactly once — the recheck must not re-plan extras the
+    // run already processed.
+    expect(ids.filter((i) => i === 'onboard.extract_ner').length).toBe(1);
+    expect(ids.filter((i) => i === 'onboard.extract_timeline').length).toBe(1);
+    expect(result.submitted.every((s) => s.status === 'completed')).toBe(true);
+    expect(submittedJobs.map((j) => j.name).sort()).toEqual([
+      'extract-ner',
+      'extract-timeline-from-meetings',
+    ]);
+  });
+});