Fix:app.py

app.py

@@ -627,11 +627,11 @@ from dotenv import load_dotenv
 from flask import Flask, render_template, request, jsonify, session
 import google.generativeai as genai
 
-# Import new langchain components and our helper
+# Import new langchain components and our helpers
 from langchain_google_genai import GoogleGenerativeAIEmbeddings
 from langchain_community.vectorstores import FAISS
 from utils.helpers import create_vector_store, get_greeting, load_quotes
-from agents.router_agent import RouterAgent
+from agents.router_agent import RouterAgent # Re-import the RouterAgent
 
 # --- Initial Setup ---
 load_dotenv()
@@ -651,19 +651,18 @@ try:
         genai.configure(api_key=GEMINI_API_KEY)
         model = genai.GenerativeModel('gemini-1.5-flash')
         
-        # Load the persistent knowledge base from disk
         if os.path.exists('faiss_index'):
             embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001")
             vector_store = FAISS.load_local("faiss_index", embeddings, allow_dangerous_deserialization=True)
             print("✅ Gemini Model and Knowledge Base loaded successfully!")
         else:
-            print("✅ Gemini Model loaded. No knowledge base found.")
+            print("✅ Gemini Model loaded. No knowledge base found to load.")
     else:
         print("⚠️ No Gemini API key found.")
 except Exception as e:
     print(f"❌ Error during initialization: {e}")
 
-# --- Main AI Application Class ---
+# --- Main AI Application Class (Reinstated) ---
 class MyPharmaAI:
     def __init__(self, gemini_model, vector_store_db):
         self.router = RouterAgent(gemini_model)
@@ -671,14 +670,14 @@ class MyPharmaAI:
         self.vector_store = vector_store_db
 
     def process_query(self, query, viva_state, chat_history):
+        # This is the core logic that combines both systems:
+        # 1. Search the permanent knowledge base for context.
         file_context = ""
         if self.vector_store:
-            # Search the knowledge base for relevant documents
-            relevant_docs = self.vector_store.similarity_search(query)
-            # Join the content of the relevant docs to create the context
+            relevant_docs = self.vector_store.similarity_search(query, k=3) # Get top 3 results
             file_context = "\n".join(doc.page_content for doc in relevant_docs)
         
-        # Pass the retrieved context to the router, maintaining the existing structure
+        # 2. Pass the retrieved context to the multi-agent router system.
         return self.router.route_query(query, file_context, viva_state, chat_history)
 
 pharma_ai = MyPharmaAI(model, vector_store)
@@ -686,32 +685,38 @@ pharma_ai = MyPharmaAI(model, vector_store)
 # --- Flask Routes ---
 @app.route('/')
 def index():
+    # Use the correct template name
     return render_template('index.html', greeting=get_greeting(), daily_quote=random.choice(pharma_ai.quotes))
 
 @app.route('/chat', methods=['POST'])
 def chat():
-    data = request.get_json()
-    query = data.get('query', '').strip()
-    if not query:
-        return jsonify({'success': False, 'error': 'Empty query'}), 400
-
-    chat_history = session.get('chat_history', [])
-    viva_state = session.get('viva_state', None)
-    
     try:
+        data = request.get_json()
+        query = data.get('query', '').strip()
+        if not query:
+            return jsonify({'success': False, 'error': 'Empty query'}), 400
+
+        chat_history = session.get('chat_history', [])
+        viva_state = session.get('viva_state', None)
+        
+        # Call the main orchestrator
         result = pharma_ai.process_query(query, viva_state, chat_history)
+
         if result.get('success'):
             chat_history.append({'role': 'user', 'parts': [query]})
             chat_history.append({'role': 'model', 'parts': [result.get('message', '')]})
             session['chat_history'] = chat_history[-10:]
+        
+        if 'viva_state' in result:
+            session['viva_state'] = result.get('viva_state')
+        
         return jsonify(result)
+        
     except Exception as e:
         print(f"Error in /chat endpoint: {e}")
         return jsonify({'success': False, 'message': f'Server error: {e}', 'agent_used': 'error'}), 500
 
 # --- Main Execution ---
 if __name__ == '__main__':
-    # Ensure data folder exists for quotes.json
-    os.makedirs('data', exist_ok=True)
     port = int(os.environ.get('PORT', 7860))
     app.run(host='0.0.0.0', port=port)