Update app/agent.py

app/agent.py

@@ -1,11 +1,53 @@
 from models.local_llm import run_llm
+import logging
 
+# Configure logging
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
+logger = logging.getLogger(__name__)
+
+# Global conversation memory (in a real app, this should be session-based)
 conversation_memory = []
 
 def process_text(input_text: str) -> str:
+    """
+    Process user input through LLM with conversation memory.
+    
+    Args:
+        input_text: User input text
+        
+    Returns:
+        Generated response
+    """
+    # Add user input to memory
     conversation_memory.append({"user": input_text})
-    context = "\n".join([f"User: {m['user']}" for m in conversation_memory])
-    prompt = f"You are a telecom AI assistant. Context:\n{context}\nRespond:"
+    
+    # Create context from memory (last 5 exchanges for brevity)
+    recent_memory = conversation_memory[-10:] if len(conversation_memory) > 10 else conversation_memory
+    context = "\n".join([
+        f"User: {m['user']}" if 'user' in m else f"Assistant: {m['assistant']}" 
+        for m in recent_memory
+    ])
+    
+    # Create prompt with context
+    prompt = f"""You are a telecom AI assistant. You help users with questions about telecommunications, 
+networks, phones, and related technologies. Be concise and helpful.
+
+Conversation history:
+{context}
+
+Please provide your next response:"""
+    
+    # Get response from LLM
+    logger.info(f"Sending prompt to LLM with context length: {len(context)}")
     response = run_llm(prompt)
+    
+    # Add response to memory
     conversation_memory.append({"assistant": response})
+    
     return response
+
+def clear_memory():
+    """Clear the conversation memory."""
+    global conversation_memory
+    conversation_memory = []
+    return True