fix: harden flaky tests and cover native google thought signatures (#23457) (thanks @echoVic)

src/agents/pi-embedded-runner.google-sanitize-thinking.e2e.test.ts

@@ -231,6 +231,72 @@ describe("sanitizeSessionHistory (google thinking)", () => {
     ]);
   });
 
+  it("strips non-base64 thought signatures for native Google Gemini", async () => {
+    const sessionManager = SessionManager.inMemory();
+    const input = [
+      {
+        role: "user",
+        content: "hi",
+      },
+      {
+        role: "assistant",
+        content: [
+          { type: "text", text: "hello", thought_signature: "msg_abc123" },
+          { type: "thinking", thinking: "ok", thought_signature: "c2ln" },
+          {
+            type: "toolCall",
+            id: "call_1",
+            name: "read",
+            arguments: { path: "/tmp/foo" },
+            thoughtSignature: '{"id":1}',
+          },
+          {
+            type: "toolCall",
+            id: "call_2",
+            name: "read",
+            arguments: { path: "/tmp/bar" },
+            thoughtSignature: "c2ln",
+          },
+        ],
+      },
+    ] as unknown as AgentMessage[];
+
+    const out = await sanitizeSessionHistory({
+      messages: input,
+      modelApi: "google-generative-ai",
+      provider: "google",
+      modelId: "gemini-2.0-flash",
+      sessionManager,
+      sessionId: "session:google-gemini",
+    });
+
+    const assistant = out.find((msg) => (msg as { role?: string }).role === "assistant") as {
+      content?: Array<{
+        type?: string;
+        thought_signature?: string;
+        thoughtSignature?: string;
+        thinking?: string;
+      }>;
+    };
+    expect(assistant.content).toEqual([
+      { type: "text", text: "hello" },
+      { type: "thinking", thinking: "ok", thought_signature: "c2ln" },
+      {
+        type: "toolCall",
+        id: "call1",
+        name: "read",
+        arguments: { path: "/tmp/foo" },
+      },
+      {
+        type: "toolCall",
+        id: "call2",
+        name: "read",
+        arguments: { path: "/tmp/bar" },
+        thoughtSignature: "c2ln",
+      },
+    ]);
+  });
+
   it("keeps mixed signed/unsigned thinking blocks for Google models", async () => {
     const sessionManager = SessionManager.inMemory();
     const input = [

src/agents/pi-embedded-runner.test.ts

@@ -130,7 +130,7 @@ beforeAll(async () => {
   workspaceDir = path.join(tempRoot, "workspace");
   await fs.mkdir(agentDir, { recursive: true });
   await fs.mkdir(workspaceDir, { recursive: true });
-}, 60_000);
+}, 180_000);
 
 afterAll(async () => {
   if (!tempRoot) {

src/agents/sessions-spawn-hooks.test.ts

@@ -1,10 +1,11 @@
-import { beforeEach, describe, expect, it, vi } from "vitest";
+import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
 import "./test-helpers/fast-core-tools.js";
 import {
   getCallGatewayMock,
   getSessionsSpawnTool,
   setSessionsSpawnConfigOverride,
 } from "./openclaw-tools.subagents.sessions-spawn.test-harness.js";
+import { resetSubagentRegistryForTests } from "./subagent-registry.js";
 
 const hookRunnerMocks = vi.hoisted(() => ({
   hasSubagentEndedHook: true,
@@ -79,6 +80,7 @@ function mockAgentStartFailure() {
 
 describe("sessions_spawn subagent lifecycle hooks", () => {
   beforeEach(() => {
+    resetSubagentRegistryForTests();
     hookRunnerMocks.hasSubagentEndedHook = true;
     hookRunnerMocks.runSubagentSpawning.mockClear();
     hookRunnerMocks.runSubagentSpawned.mockClear();
@@ -103,6 +105,10 @@ describe("sessions_spawn subagent lifecycle hooks", () => {
     });
   });
 
+  afterEach(() => {
+    resetSubagentRegistryForTests();
+  });
+
   it("runs subagent_spawning and emits subagent_spawned with requester metadata", async () => {
     const tool = await getSessionsSpawnTool({
       agentSessionKey: "main",

src/agents/transcript-policy.test.ts

@@ -19,6 +19,10 @@ describe("resolveTranscriptPolicy", () => {
       modelApi: "google-generative-ai",
     });
     expect(policy.sanitizeToolCallIds).toBe(true);
+    expect(policy.sanitizeThoughtSignatures).toEqual({
+      allowBase64Only: true,
+      includeCamelCase: true,
+    });
   });
 
   it("enables sanitizeToolCallIds for Mistral provider", () => {