From 2506c4728ee00f072411b73587bda24b46e8e565 Mon Sep 17 00:00:00 2001
From: garrytan <garrytan@gmail.com>
Date: Tue, 29 Sep 2026 15:32:07 +0000
Subject: [PATCH] test: start the eng batching eval with its setup
 prerequisites supplied

Routing setup and cross-project learnings (D1/D2 in run 36385945043) are
never counted and are not what the case measures. The registration now uses
the runner's existing preconfiguredReviewActor so the attempt starts at the
review; engSetupAUQ still vetoes any late setup question. The registration
test pins the option.
---
 test/eng-batching-saved-ledger.test.ts                 | 6 +++---
 test/skill-e2e-plan-eng-multi-finding-batching.test.ts | 6 +++++-
 2 files changed, 8 insertions(+), 4 deletions(-)

diff --git a/test/eng-batching-saved-ledger.test.ts b/test/eng-batching-saved-ledger.test.ts
index d9dd7eb57..e7f3d2210 100644
--- a/test/eng-batching-saved-ledger.test.ts
+++ b/test/eng-batching-saved-ledger.test.ts
@@ -451,14 +451,14 @@ test.each([
 import { describe, expect, mock } from 'bun:test';
 import * as fs from 'node:fs';
 import * as real from ${JSON.stringify(runner)};
-const facts = { runs: 0, stops: [] as boolean[], ceiling: 0 };
+const facts = { runs: 0, stops: [] as boolean[], ceiling: 0, preconfigured: false };
 const save = () => fs.writeFileSync(${JSON.stringify(factsPath)}, JSON.stringify(facts));
 const fp = (signature: string, preReview: boolean, administrative?: string) => ({ signature, preReview, administrative, promptSnippet: signature, options: [], observedAtMs: 1 });
 mock.module(${JSON.stringify(path.join(ROOT, 'test/helpers/e2e-gate.ts'))}, () => ({
   describeE2ETier: (tier: string) => { expect(tier).toBe('periodic'); return describe; },
 }));
 mock.module(${JSON.stringify(runner)}, () => ({ ...real, runPlanSkillCounting: async (opts: any) => {
-  facts.runs++; facts.ceiling = opts.reviewCountCeiling;
+  facts.runs++; facts.ceiling = opts.reviewCountCeiling; facts.preconfigured = opts.preconfiguredReviewActor;
   const setup = [fp('s1', true), fp('s2', true)];
   const review = [fp('r1', false), fp('r2', false), fp('r3', false)];
   facts.stops = [
@@ -479,7 +479,7 @@ await import(${JSON.stringify(path.join(ROOT, 'test/skill-e2e-plan-eng-multi-fin
     const [exit, out, err] = await Promise.all([child.exited, new Response(child.stdout).text(), new Response(child.stderr).text()]);
     const facts = JSON.parse(fs.readFileSync(factsPath, 'utf8'));
     expect(exit, out + err).toBe(passes ? 0 : 1);
-    expect(facts).toEqual({ runs: 1, stops: [false, false, true], ceiling: 7 });
+    expect(facts).toEqual({ runs: 1, stops: [false, false, true], ceiling: 7, preconfigured: true });
   } finally {
     fs.rmSync(temp, { recursive: true, force: true });
   }
diff --git a/test/skill-e2e-plan-eng-multi-finding-batching.test.ts b/test/skill-e2e-plan-eng-multi-finding-batching.test.ts
index 424b3d6f1..7c94803f1 100644
--- a/test/skill-e2e-plan-eng-multi-finding-batching.test.ts
+++ b/test/skill-e2e-plan-eng-multi-finding-batching.test.ts
@@ -22,7 +22,7 @@
  *     This is the tightest regression test for the original bug class —
  *     not a band-around-N test, but a "did the agent batch?" test.
  *
- * Tier: periodic (~7 min observed; 25 min budget). Sequential by default.
+ * Tier: periodic (~6 min expected; 25 min budget). Sequential by default.
  */
 
 import { test } from 'bun:test';
@@ -88,6 +88,10 @@ describeE2E('/plan-eng-review multi-finding batching regression (periodic)', ()
           isCollectionComplete: (_transcript, fingerprints) =>
             fingerprints.filter(fp => !fp.preReview && !fp.administrative).length >= FLOOR,
           reviewCountCeiling: N + 3, // hard cap above floor + tolerance
+          // Supplied prerequisites: routing setup and cross-project learnings are
+          // already declined, so the attempt starts at the review (setup answers
+          // were never counted; engSetupAUQ still vetoes any late setup question).
+          preconfiguredReviewActor: true,
           timeoutMs: 1_500_000, // 25 min
           env: { QUESTION_TUNING: 'false', EXPLAIN_LEVEL: 'default' },
         });