Fix Gemini model name for API compatibility

- Use models/gemini-2.0-flash-001 as the correct model path
- Simplify to generate_content instead of chat API
- Include conversation history in prompt context

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -844,20 +844,21 @@ def chat_with_gemini(
 
     try:
         genai.configure(api_key=api_key)
-        model = genai.GenerativeModel('gemini-1.5-flash')
 
-        # Build conversation history for context
-        chat_history = []
-        for msg in history:
-            role = "user" if msg.get("role") == "user" else "model"
-            chat_history.append({"role": role, "parts": [msg.get("content", "")]})
+        # Use models/gemini-2.0-flash-001 which is the correct model path
+        model = genai.GenerativeModel('models/gemini-2.0-flash-001')
 
-        # Start chat with history
-        chat = model.start_chat(history=chat_history)
+        # Build context from history
+        context_parts = [LINEAGE_AGENT_PROMPT, "\n\nConversation history:"]
+        for msg in history[-6:]:  # Keep last 6 messages for context
+            role = "User" if msg.get("role") == "user" else "Assistant"
+            context_parts.append(f"{role}: {msg.get('content', '')}")
 
-        # Send message with system prompt context
-        full_prompt = f"{LINEAGE_AGENT_PROMPT}\n\nUser query: {message}"
-        response = chat.send_message(full_prompt)
+        context_parts.append(f"\nUser: {message}\nAssistant:")
+
+        # Generate response
+        full_prompt = "\n".join(context_parts)
+        response = model.generate_content(full_prompt)
 
         assistant_message = response.text