Merge branch 'pr-1368' into integration/validation-batch

# Conflicts:
#	plugin/scripts/context-generator.cjs
#	plugin/scripts/mcp-server.cjs
#	plugin/scripts/worker-service.cjs
#	plugin/ui/viewer-bundle.js

tests/gemini_agent.test.ts

@@ -358,6 +358,90 @@ describe('GeminiAgent', () => {
     }
   });
 
+  describe('conversation history truncation', () => {
+    it('should truncate history when message count exceeds limit', async () => {
+      // Build a history with 25 small messages (limit is 20)
+      const history: any[] = [];
+      for (let i = 0; i < 25; i++) {
+        history.push({ role: i % 2 === 0 ? 'user' : 'assistant', content: `message ${i}` });
+      }
+
+      const session = {
+        sessionDbId: 1,
+        contentSessionId: 'test-session',
+        memorySessionId: 'mem-session-123',
+        project: 'test-project',
+        userPrompt: 'test prompt',
+        conversationHistory: history,
+        lastPromptNumber: 2,
+        cumulativeInputTokens: 0,
+        cumulativeOutputTokens: 0,
+        pendingMessages: [],
+        abortController: new AbortController(),
+        generatorPromise: null,
+        earliestPendingTimestamp: null,
+        currentProvider: null,
+        startTime: Date.now(),
+        processingMessageIds: []
+      } as any;
+
+      global.fetch = mock(() => Promise.resolve(new Response(JSON.stringify({
+        candidates: [{ content: { parts: [{ text: 'response' }] } }]
+      }))));
+
+      await agent.startSession(session);
+
+      // The request body should have truncated contents (init adds 1 more, so 26 total → truncated to 20)
+      const body = JSON.parse((global.fetch as any).mock.calls[0][1].body);
+      expect(body.contents.length).toBeLessThanOrEqual(20);
+    });
+
+    it('should always keep at least the newest message even if it exceeds token limit', async () => {
+      // Override settings to have a very low token limit
+      loadFromFileSpy.mockImplementation(() => ({
+        ...SettingsDefaultsManager.getAllDefaults(),
+        CLAUDE_MEM_GEMINI_API_KEY: 'test-api-key',
+        CLAUDE_MEM_GEMINI_MODEL: 'gemini-2.5-flash-lite',
+        CLAUDE_MEM_GEMINI_RATE_LIMITING_ENABLED: 'false',
+        CLAUDE_MEM_GEMINI_MAX_CONTEXT_MESSAGES: '20',
+        CLAUDE_MEM_GEMINI_MAX_TOKENS: '1000',  // Very low: ~250 chars
+        CLAUDE_MEM_DATA_DIR: '/tmp/claude-mem-test',
+      }));
+
+      // Create a single large message that exceeds the token limit
+      const largeContent = 'x'.repeat(8000);  // ~2000 tokens, well above 1000 limit
+
+      const session = {
+        sessionDbId: 1,
+        contentSessionId: 'test-session',
+        memorySessionId: 'mem-session-123',
+        project: 'test-project',
+        userPrompt: largeContent,
+        conversationHistory: [],
+        lastPromptNumber: 1,
+        cumulativeInputTokens: 0,
+        cumulativeOutputTokens: 0,
+        pendingMessages: [],
+        abortController: new AbortController(),
+        generatorPromise: null,
+        earliestPendingTimestamp: null,
+        currentProvider: null,
+        startTime: Date.now(),
+        processingMessageIds: []
+      } as any;
+
+      global.fetch = mock(() => Promise.resolve(new Response(JSON.stringify({
+        candidates: [{ content: { parts: [{ text: 'response' }] } }]
+      }))));
+
+      await agent.startSession(session);
+
+      // Should still send at least 1 message (the newest), not empty contents
+      const body = JSON.parse((global.fetch as any).mock.calls[0][1].body);
+      expect(body.contents.length).toBeGreaterThanOrEqual(1);
+    });
+  });
+
   describe('gemini-3-flash-preview model support', () => {
     it('should accept gemini-3-flash-preview as a valid model', async () => {
       // The GeminiModel type includes gemini-3-flash-preview - compile-time check