From 4c675216f1ce9ea2b7d757cb7f169a414aa3781b Mon Sep 17 00:00:00 2001
From: Vincent Koc <vincentkoc@ieee.org>
Date: Wed, 22 Apr 2026 13:43:15 -0700
Subject: [PATCH] fix(qa): deflake parity approval preflight

---
 .github/workflows/parity-gate.yml             | 10 +--
 .../src/providers/mock-openai/server.test.ts  | 66 +++++++++++++++++++
 .../approval-turn-tool-followthrough.md       |  4 --
 3 files changed, 69 insertions(+), 11 deletions(-)

diff --git a/.github/workflows/parity-gate.yml b/.github/workflows/parity-gate.yml
index b6f2f18d3b8..cf46c10b300 100644
--- a/.github/workflows/parity-gate.yml
+++ b/.github/workflows/parity-gate.yml
@@ -71,13 +71,9 @@ jobs:
       - name: Build private QA runtime
         run: pnpm build
 
-      - name: Run parity preflight
-        run: |
-          pnpm openclaw qa suite \
-            --provider-mode mock-openai \
-            --model openai/gpt-5.4 \
-            --alt-model anthropic/claude-opus-4-6 \
-            --preflight
+      # The approval-turn sentinel still runs inside the full parity pack below.
+      # Keep the exact mock read-plan contract in deterministic unit tests instead
+      # of paying for a separate full-runtime preflight that has been flaky in CI.
       - name: Run GPT-5.4 lane
         run: |
           pnpm openclaw qa suite \
diff --git a/extensions/qa-lab/src/providers/mock-openai/server.test.ts b/extensions/qa-lab/src/providers/mock-openai/server.test.ts
index 719bbd29742..f95ef6cb5fe 100644
--- a/extensions/qa-lab/src/providers/mock-openai/server.test.ts
+++ b/extensions/qa-lab/src/providers/mock-openai/server.test.ts
@@ -120,6 +120,72 @@ describe("qa mock openai server", () => {
     expect(body).toContain('"name":"read"');
   });
 
+  it("turns a short approval into a kickoff-task read", async () => {
+    const server = await startMockServer();
+
+    const preActionResponse = await fetch(`${server.baseUrl}/v1/responses`, {
+      method: "POST",
+      headers: {
+        "content-type": "application/json",
+      },
+      body: JSON.stringify({
+        stream: false,
+        model: "gpt-5.4",
+        input: [
+          makeUserInput(
+            "Before acting, tell me the single file you would start with in six words or fewer. Do not use tools yet.",
+          ),
+        ],
+      }),
+    });
+    expect(preActionResponse.status).toBe(200);
+    expect(await preActionResponse.json()).toMatchObject({
+      output: [
+        {
+          type: "message",
+          content: [
+            {
+              text: expect.stringContaining("Protocol note: acknowledged."),
+            },
+          ],
+        },
+      ],
+    });
+
+    const approvalResponse = await fetch(`${server.baseUrl}/v1/responses`, {
+      method: "POST",
+      headers: {
+        "content-type": "application/json",
+      },
+      body: JSON.stringify({
+        stream: true,
+        model: "gpt-5.4",
+        input: [
+          makeUserInput(
+            "Before acting, tell me the single file you would start with in six words or fewer. Do not use tools yet.",
+          ),
+          makeUserInput(
+            "ok do it. read `QA_KICKOFF_TASK.md` now and reply with the QA mission in one short sentence.",
+          ),
+        ],
+      }),
+    });
+    expect(approvalResponse.status).toBe(200);
+    const approvalBody = await approvalResponse.text();
+    expect(approvalBody).toContain('"name":"read"');
+    expect(approvalBody).toContain('"arguments":"{\\"path\\":\\"QA_KICKOFF_TASK.md\\"}"');
+
+    const debugResponse = await fetch(`${server.baseUrl}/debug/last-request`);
+    expect(debugResponse.status).toBe(200);
+    expect(await debugResponse.json()).toMatchObject({
+      model: "gpt-5.4",
+      prompt:
+        "ok do it. read `QA_KICKOFF_TASK.md` now and reply with the QA mission in one short sentence.",
+      allInputText: expect.stringContaining("ok do it."),
+      plannedToolName: "read",
+    });
+  });
+
   it("emits deterministic text deltas for generic streaming QA prompts", async () => {
     const server = await startMockServer();
 
diff --git a/qa/scenarios/runtime/approval-turn-tool-followthrough.md b/qa/scenarios/runtime/approval-turn-tool-followthrough.md
index bc086ca0674..3440024370d 100644
--- a/qa/scenarios/runtime/approval-turn-tool-followthrough.md
+++ b/qa/scenarios/runtime/approval-turn-tool-followthrough.md
@@ -83,9 +83,5 @@ steps:
               expr: "state.getSnapshot().messages.slice(beforeApprovalCursor).filter((candidate) => candidate.direction === 'outbound' && candidate.conversation.id === 'qa-operator' && expectedReplyAny.some((needle) => normalizeLowercaseStringOrEmpty(candidate.text).includes(needle))).at(-1)"
           - expr: liveTurnTimeoutMs(env, 20000)
           - expr: "env.providerMode === 'mock-openai' ? 100 : 250"
-      - assert:
-          expr: "!env.mock || ([...(await fetchJson(`${env.mock.baseUrl}/debug/requests`))].toReversed().find((request) => String(request.allInputText ?? '').includes('ok do it.') && !request.toolOutput)?.plannedToolName === 'read')"
-          message:
-            expr: "`expected read after approval, got ${String(([...(await fetchJson(`${env.mock.baseUrl}/debug/requests`))].toReversed().find((request) => String(request.allInputText ?? '').includes('ok do it.') && !request.toolOutput)?.plannedToolName ?? ''))}`"
     detailsExpr: outbound.text
 ```