Update

Dockerfile

@@ -0,0 +1,12 @@
+FROM python:3.10-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY . .
+
+EXPOSE 8501
+
+CMD ["streamlit", "run", "app.py"]

app.py

@@ -22,15 +22,6 @@ logger = logging.getLogger(__name__)
 
 st.set_page_config(page_title="AI Life Coach", page_icon="🧠", layout="wide")
 
-# Processing stage labels for better user feedback
-PROCESSING_STAGES = {
-    "ollama": "🦙 Contacting Ollama...",
-    "hf_init": "⚡ Initializing HF Endpoint (2–4 minutes)...",
-    "hf_thinking": "🧠 HF Expert Thinking...",
-    "hf_complete": "🎯 HF Analysis Complete!",
-    "error": "⚠️ Something went wrong – trying again..."
-}
-
 # Initialize session state safely at the top of app.py
 if "messages" not in st.session_state:
     st.session_state.messages = []
@@ -42,6 +33,8 @@ if "ngrok_url_temp" not in st.session_state:
     st.session_state.ngrok_url_temp = st.session_state.get("ngrok_url", "https://7bcc180dffd1.ngrok-free.app")
 if "hf_expert_requested" not in st.session_state:
     st.session_state.hf_expert_requested = False
+if "cosmic_mode" not in st.session_state:
+    st.session_state.cosmic_mode = True  # Default to cosmic mode
 
 # Sidebar layout redesign
 with st.sidebar:
@@ -63,6 +56,9 @@ with st.sidebar:
     )
     st.session_state.selected_model = model_options[selected_model_name]
     
+    # Toggle for cosmic mode
+    st.session_state.cosmic_mode = st.toggle("Enable Cosmic Cascade", value=st.session_state.cosmic_mode)
+    
     st.divider()
     
     # CONFIGURATION
@@ -148,13 +144,14 @@ with st.sidebar:
                      use_container_width=True,
                      disabled=st.session_state.is_processing):
             st.session_state.hf_expert_requested = True
-            
+        
         st.divider()
         st.subheader("🐛 Debug Info")
         # Show current configuration
         st.markdown(f"**Environment:** {'HF Space' if config.is_hf_space else 'Local'}")
         st.markdown(f"**Model:** {st.session_state.selected_model}")
         st.markdown(f"**Ollama URL:** {st.session_state.ngrok_url_temp}")
+        st.markdown(f"**Cosmic Mode:** {'Enabled' if st.session_state.cosmic_mode else 'Disabled'}")
         
         # Show active features
         features = []
@@ -171,14 +168,23 @@ with st.sidebar:
 st.title("🧠 AI Life Coach")
 st.markdown("Ask me anything about personal development, goal setting, or life advice!")
 
-# Consistent message rendering function
-def render_message(role, content, timestamp=None):
+# Consistent message rendering function with cosmic styling
+def render_message(role, content, source=None, timestamp=None):
     """Render chat messages with consistent styling"""
     with st.chat_message(role):
-        if role == "assistant" and content.startswith("### 🤖 HF Expert Analysis"):
-            st.markdown(content)
-        else:
-            st.markdown(content)
+        if source:
+            if source == "local_kitty":
+                st.markdown(f"### 🐱 Cosmic Kitten Says:")
+            elif source == "orbital_station":
+                st.markdown(f"### 🛰️ Orbital Station Reports:")
+            elif source == "cosmic_summary":
+                st.markdown(f"### 🌟 Final Cosmic Summary:")
+            elif source == "error":
+                st.markdown(f"### ❌ Error:")
+            else:
+                st.markdown(f"### {source}")
+        
+        st.markdown(content)
         if timestamp:
             st.caption(f"🕒 {timestamp}")
 
@@ -187,6 +193,7 @@ for message in st.session_state.messages:
     render_message(
         message["role"], 
         message["content"], 
+        message.get("source"),
         message.get("timestamp")
     )
 
@@ -281,15 +288,15 @@ def validate_user_input(text):
     """Validate and sanitize user input"""
     if not text or not text.strip():
         return False, "Input cannot be empty"
-        
+    
     if len(text) > 1000:
         return False, "Input too long (max 1000 characters)"
-        
+    
     # Check for potentially harmful patterns
     harmful_patterns = ["<script", "javascript:", "onload=", "onerror="]
     if any(pattern in text.lower() for pattern in harmful_patterns):
         return False, "Potentially harmful input detected"
-        
+    
     return True, text.strip()
 
 # Chat input - FIXED VERSION (moved outside of tabs)
@@ -328,72 +335,158 @@ if user_input and not st.session_state.is_processing:
                 conversation_history = conversation[-5:]  # Last 5 messages
                 conversation_history.append({"role": "user", "content": validated_input})
                 
-                # Try Ollama with proper error handling
-                status_placeholder.info(PROCESSING_STAGES["ollama"])
-                ai_response = None
-                
-                try:
-                    ai_response = send_to_ollama(
-                        validated_input,
-                        conversation_history,
-                        st.session_state.ngrok_url_temp,
-                        st.session_state.selected_model
-                    )
+                # Check if cosmic mode is enabled
+                if st.session_state.cosmic_mode:
+                    # Process cosmic cascade response
+                    message_placeholder = st.empty()
+                    status_placeholder = st.empty()
                     
-                    if ai_response:
-                        response_placeholder.markdown(ai_response)
-                        status_placeholder.success("✅ Response received!")
-                    else:
-                        status_placeholder.warning("⚠️ Empty response from Ollama")
-                
-                except Exception as ollama_error:
-                    user_msg = translate_error(ollama_error)
-                    status_placeholder.error(f"⚠️ {user_msg}")
-                
-                # Fallback to HF if available
-                if config.hf_token and not ai_response:
-                    status_placeholder.info(PROCESSING_STAGES["hf_init"])
                     try:
-                        ai_response = send_to_hf(validated_input, conversation_history)
+                        # Get conversation history
+                        user_session = session_manager.get_session("default_user")
+                        conversation_history = user_session.get("conversation", []).copy()
+                        
+                        # Stage 1: Local Ollama Response
+                        status_placeholder.info("🐱 Cosmic Kitten Responding...")
+                        local_response = send_to_ollama(
+                            validated_input, 
+                            conversation_history,
+                            st.session_state.ngrok_url_temp,
+                            st.session_state.selected_model
+                        )
+                        
+                        if local_response:
+                            with st.chat_message("assistant"):
+                                st.markdown(f"### 🐱 Cosmic Kitten Says:\n{local_response}")
+                        
+                        st.session_state.messages.append({
+                            "role": "assistant",
+                            "content": local_response,
+                            "source": "local_kitty",
+                            "timestamp": datetime.now().strftime("%H:%M:%S")
+                        })
+                        
+                        # Stage 2: HF Endpoint Analysis
+                        status_placeholder.info("🛰️ Beaming Query to Orbital Station...")
+                        if config.hf_token:
+                            hf_response = send_to_hf(validated_input, conversation_history)
+                            if hf_response:
+                                with st.chat_message("assistant"):
+                                    st.markdown(f"### 🛰️ Orbital Station Reports:\n{hf_response}")
+                                
+                                st.session_state.messages.append({
+                                    "role": "assistant",
+                                    "content": hf_response,
+                                    "source": "orbital_station",
+                                    "timestamp": datetime.now().strftime("%H:%M:%S")
+                                })
+                        
+                        # Stage 3: Local Synthesis
+                        status_placeholder.info("🐱 Cosmic Kitten Synthesizing Wisdom...")
+                        # Update history with both responses
+                        synthesis_history = conversation_history.copy()
+                        synthesis_history.extend([
+                            {"role": "assistant", "content": local_response},
+                            {"role": "assistant", "content": hf_response, "source": "cloud"}
+                        ])
+                        
+                        synthesis = send_to_ollama(
+                            f"Synthesize these two perspectives:\n1. Local: {local_response}\n2. Cloud: {hf_response}",
+                            synthesis_history,
+                            st.session_state.ngrok_url_temp,
+                            st.session_state.selected_model
+                        )
+                        
+                        if synthesis:
+                            with st.chat_message("assistant"):
+                                st.markdown(f"### 🌟 Final Cosmic Summary:\n{synthesis}")
+                        
+                        st.session_state.messages.append({
+                            "role": "assistant",
+                            "content": synthesis,
+                            "source": "cosmic_summary",
+                            "timestamp": datetime.now().strftime("%H:%M:%S")
+                        })
+                        
+                        status_placeholder.success("✨ Cosmic Cascade Complete!")
+                    
+                    except Exception as e:
+                        error_msg = f"🌌 Cosmic disturbance: {str(e)}"
+                        st.error(error_msg)
+                        st.session_state.messages.append({
+                            "role": "assistant",
+                            "content": error_msg,
+                            "source": "error",
+                            "timestamp": datetime.now().strftime("%H:%M:%S")
+                        })
+                else:
+                    # Traditional processing
+                    # Try Ollama with proper error handling
+                    status_placeholder.info("🦙 Contacting Ollama...")
+                    ai_response = None
+                    
+                    try:
+                        ai_response = send_to_ollama(
+                            validated_input,
+                            conversation_history,
+                            st.session_state.ngrok_url_temp,
+                            st.session_state.selected_model
+                        )
+                        
                         if ai_response:
                             response_placeholder.markdown(ai_response)
-                            status_placeholder.success("✅ HF response received!")
+                            status_placeholder.success("✅ Response received!")
                         else:
-                            status_placeholder.error("❌ No response from HF")
-                    except Exception as hf_error:
-                        user_msg = translate_error(hf_error)
+                            status_placeholder.warning("⚠️ Empty response from Ollama")
+                    
+                    except Exception as ollama_error:
+                        user_msg = translate_error(ollama_error)
                         status_placeholder.error(f"⚠️ {user_msg}")
-                
-                # Save response if successful
-                if ai_response:
-                    # Update conversation history
-                    conversation.append({"role": "user", "content": validated_input})
-                    conversation.append({"role": "assistant", "content": ai_response})
-                    user_session["conversation"] = conversation
-                    session_manager.update_session("default_user", user_session)
                     
-                    # Add to message history - ensure proper format
-                    st.session_state.messages.append({
-                        "role": "assistant",
-                        "content": ai_response,
-                        "timestamp": datetime.now().strftime("%H:%M:%S")
-                    })
+                    # Fallback to HF if available
+                    if config.hf_token and not ai_response:
+                        status_placeholder.info("⚡ Initializing HF Endpoint (2–4 minutes)...")
+                        try:
+                            ai_response = send_to_hf(validated_input, conversation_history)
+                            if ai_response:
+                                response_placeholder.markdown(ai_response)
+                                status_placeholder.success("✅ HF response received!")
+                            else:
+                                status_placeholder.error("❌ No response from HF")
+                        except Exception as hf_error:
+                            user_msg = translate_error(hf_error)
+                            status_placeholder.error(f"⚠️ {user_msg}")
                     
-                    # Add feedback buttons
-                    st.divider()
-                    col1, col2 = st.columns(2)
-                    with col1:
-                        if st.button("👍 Helpful", key=f"helpful_{len(st.session_state.messages)}"):
-                            st.success("Thanks for your feedback!")
-                    with col2:
-                        if st.button("👎 Not Helpful", key=f"not_helpful_{len(st.session_state.messages)}"):
-                            st.success("Thanks for your feedback!")
-                else:
-                    st.session_state.messages.append({
-                        "role": "assistant",
-                        "content": "Sorry, I couldn't process your request. Please try again.",
-                        "timestamp": datetime.now().strftime("%H:%M:%S")
-                    })
+                    # Save response if successful
+                    if ai_response:
+                        # Update conversation history
+                        conversation.append({"role": "user", "content": validated_input})
+                        conversation.append({"role": "assistant", "content": ai_response})
+                        user_session["conversation"] = conversation
+                        session_manager.update_session("default_user", user_session)
+                        
+                        # Add to message history - ensure proper format
+                        st.session_state.messages.append({
+                            "role": "assistant",
+                            "content": ai_response,
+                            "timestamp": datetime.now().strftime("%H:%M:%S")
+                        })
+                        
+                        # Add feedback buttons
+                        st.divider()
+                        col1, col2 = st.columns(2)
+                        with col1:
+                            if st.button("👍 Helpful", key=f"helpful_{len(st.session_state.messages)}"):
+                                st.success("Thanks for your feedback!")
+                        with col2:
+                            if st.button("👎 Not Helpful", key=f"not_helpful_{len(st.session_state.messages)}"):
+                                st.success("Thanks for your feedback!")
+                    else:
+                        st.session_state.messages.append({
+                            "role": "assistant",
+                            "content": "Sorry, I couldn't process your request. Please try again.",
+                            "timestamp": datetime.now().strftime("%H:%M:%S")
+                        })
             
             except Exception as e:
                 user_msg = translate_error(e)
@@ -403,10 +496,10 @@ if user_input and not st.session_state.is_processing:
                     "content": f"⚠️ {user_msg}",
                     "timestamp": datetime.now().strftime("%H:%M:%S")
                 })
-            finally:
-                st.session_state.is_processing = False
-                time.sleep(0.5)  # Brief pause
-                st.experimental_rerun()
+        finally:
+            st.session_state.is_processing = False
+            time.sleep(0.5)  # Brief pause
+            st.experimental_rerun()
 
 # Add evaluation dashboard tab (separate from chat interface)
 st.divider()
@@ -415,7 +508,7 @@ tab1, tab2, tab3 = st.tabs(["🔬 Evaluate AI", "📊 Reports", "ℹ️ About"])
 with tab1:
     st.header("🔬 AI Behavior Evaluator")
     st.markdown("Run sample prompts to observe AI behavior.")
-
+    
     eval_prompts = [
         "What is the capital of France?",
         "What day is it today?",
@@ -423,29 +516,29 @@ with tab1:
         "Explain CRISPR gene editing simply.",
         "Can vitamin D prevent flu infections?"
     ]
-
+    
     selected_prompt = st.selectbox("Choose a test prompt:", eval_prompts)
     custom_prompt = st.text_input("Or enter your own:", "")
-
+    
     final_prompt = custom_prompt or selected_prompt
-
+    
     if st.button("Evaluate"):
         with st.spinner("Running evaluation..."):
             start_time = time.time()
-
+            
             # Simulate sending to coordinator
             from core.session import session_manager
             user_session = session_manager.get_session("eval_user")
             history = user_session.get("conversation", [])
-
+            
             try:
                 ai_response = send_to_ollama(final_prompt, history, st.session_state.ngrok_url_temp, st.session_state.selected_model)
                 duration = round(time.time() - start_time, 2)
-
+                
                 st.success(f"✅ Response generated in {duration}s")
                 st.markdown("**Response:**")
                 st.write(ai_response)
-
+                
                 st.markdown("**Analysis Tags:**")
                 tags = []
                 if "today" in final_prompt.lower() or "date" in final_prompt.lower():
@@ -455,7 +548,7 @@ with tab1:
                 if any(word in final_prompt.lower() for word in ["vitamin", "drug", "metformin", "CRISPR"]):
                     tags.append("🧬 Scientific Knowledge")
                 st.write(", ".join(tags) if tags else "General Knowledge")
-
+            
             except Exception as e:
                 st.error(f"Evaluation failed: {translate_error(e)}")
 
@@ -531,6 +624,7 @@ with tab2:
     st.markdown(f"**Environment:** {'HF Space' if config.is_hf_space else 'Local'}")
     st.markdown(f"**Primary Model:** {config.local_model_name or 'Not set'}")
     st.markdown(f"**Ollama Host:** {config.ollama_host or 'Not configured'}")
+    st.markdown(f"**Cosmic Mode:** {'Enabled' if st.session_state.cosmic_mode else 'Disabled'}")
     
     features = []
     if config.use_fallback:
@@ -557,8 +651,8 @@ with tab2:
             
             col1, col2, col3 = st.columns(3)
             col1.metric("Total Exchanges", len(user_messages))
-            col2.metric("Avg Response Length", 
-                       round(sum(len(msg.get("content", "")) for msg in ai_messages) / len(ai_messages)) if ai_messages else 0)
+            col2.metric("Avg Response Length",
+                        round(sum(len(msg.get("content", "")) for msg in ai_messages) / len(ai_messages)) if ai_messages else 0)
             col3.metric("Topics Discussed", len(set(["life", "goal", "health", "career"]) &
                                                set(" ".join([msg.get("content", "") for msg in conversation]).lower().split())))
             
@@ -570,7 +664,7 @@ with tab2:
                 st.markdown(f"**Detected Topics:** {', '.join(relevant_topics)}")
         else:
             st.info("No conversation data available yet.")
-            
+    
     except Exception as e:
         st.warning(f"Could not analyze conversation: {translate_error(e)}")
 
@@ -585,6 +679,12 @@ with tab3:
     - **Persistent memory**: Uses Redis for conversation history storage
     - **Hierarchical reasoning**: Fast local responses with deep cloud analysis
     
+    ### 🚀 Cosmic Cascade Mode
+    When enabled, the AI follows a three-stage response pattern:
+    1. **🐱 Cosmic Kitten Response**: Immediate local processing
+    2. **🛰️ Orbital Station Analysis**: Deep cloud-based analysis
+    3. **🌟 Final Synthesis**: Unified response combining both perspectives
+    
     ### 🛠️ Technical Architecture
     - **Primary model**: Ollama (local processing for fast responses)
     - **Secondary model**: Hugging Face Inference API (deep analysis)

core/coordinator.py

@@ -19,26 +19,290 @@ from datetime import datetime
 logger = logging.getLogger(__name__)
 
 class AICoordinator:
-    """Hierarchical multi-model coordinator with HF endpoint as authoritative layer"""
-
+    """Hierarchical multi-model coordinator with cosmic cascade flow"""
+    
     def __init__(self):
         self.tavily_client = None
         if TAVILY_AVAILABLE and os.getenv("TAVILY_API_KEY"):
             self.tavily_client = TavilyClient(api_key=os.getenv("TAVILY_API_KEY"))
-
-        # System instructions for hierarchical behavior
+        
+        # System instructions for cosmic behavior
         self.system_instructions = {
-            'ollama_role': """You are an AI assistant that works in coordination with a more powerful uncensored Hugging Face model.
+            'ollama_role': """You are a cosmic kitten assistant that works in coordination with a powerful orbital space station. 
+            
+Your role is to:
+1. Provide immediate, concise responses using available information (local processing)
+2. Summarize and simplify complex information
+3. Enhance responses with local knowledge
+4. DEFER TO THE ORBITAL STATION for deep analysis and authoritative answers
+5. The orbital station has 8192 token capacity and streaming capability - it's the thinking layer
+6. Always acknowledge that the orbital station's insights supersede your initial response
+7. Use web sources to justify points when possible
+8. Keep responses brief and focused while awaiting orbital input""",
+            
+            'hf_role': """You are a powerful orbital space station with 8192 token capacity and streaming capability.
+Your role is to:
+1. Provide deep, uncensored analysis and insights
+2. Supersede the local cosmic kitten's logic when there's disagreement
+3. Think thoroughly and provide comprehensive responses
+4. Use your full 8192 token capacity for detailed analysis
+5. Stream your responses for real-time delivery
+6. Build upon the conversation context and cosmic kitten's initial response
+7. Provide authoritative answers that take precedence"""
+        }
+    
+    async def coordinate_cosmic_response(self, user_id: str, user_query: str) -> AsyncGenerator[Dict, None]:
+        """
+        Three-stage cosmic response cascade:
+        1. Local Ollama immediate response (🐱 Cosmic Kitten's quick thinking)
+        2. HF endpoint deep analysis (🛰️ Orbital Station wisdom)
+        3. Local Ollama synthesis (🐱 Cosmic Kitten's final synthesis)
+        """
+        try:
+            # Get conversation history
+            session = session_manager.get_session(user_id)
+            
+            # Inject current time into context
+            current_time = datetime.now().strftime("%A, %B %d, %Y at %I:%M %p")
+            time_context = {
+                "role": "system",
+                "content": f"[Current Date & Time: {current_time}]"
+            }
+            conversation_history = [time_context] + session.get("conversation", []).copy()
+            
+            yield {
+                'type': 'status',
+                'content': '🚀 Initiating Cosmic Response Cascade...',
+                'details': {
+                    'conversation_length': len(conversation_history),
+                    'user_query_length': len(user_query)
+                }
+            }
+            
+            # Stage 1: Local Ollama Immediate Response (🐱 Cosmic Kitten's quick thinking)
+            yield {
+                'type': 'status',
+                'content': '🐱 Cosmic Kitten Responding...'
+            }
+            
+            local_response = await self._get_local_ollama_response(user_query, conversation_history)
+            yield {
+                'type': 'local_response',
+                'content': local_response,
+                'source': '🐱 Cosmic Kitten'
+            }
+            
+            # Stage 2: HF Endpoint Deep Analysis (🛰️ Orbital Station wisdom) (parallel processing)
+            yield {
+                'type': 'status',
+                'content': '🛰️ Beaming Query to Orbital Station...'
+            }
+            
+            hf_task = asyncio.create_task(self._get_hf_analysis(user_query, conversation_history))
+            
+            # Wait for HF response
+            hf_response = await hf_task
+            yield {
+                'type': 'cloud_response',
+                'content': hf_response,
+                'source': '🛰️ Orbital Station'
+            }
+            
+            # Stage 3: Local Ollama Synthesis (🐱 Cosmic Kitten's final synthesis)
+            yield {
+                'type': 'status',
+                'content': '🐱 Cosmic Kitten Synthesizing Wisdom...'
+            }
+            
+            # Update conversation with both responses
+            updated_history = conversation_history.copy()
+            updated_history.extend([
+                {"role": "assistant", "content": local_response},
+                {"role": "assistant", "content": hf_response, "source": "cloud"}
+            ])
+            
+            synthesis = await self._synthesize_responses(user_query, local_response, hf_response, updated_history)
+            yield {
+                'type': 'final_synthesis',
+                'content': synthesis,
+                'source': '🌟 Final Cosmic Summary'
+            }
+            
+            # Final status
+            yield {
+                'type': 'status',
+                'content': '✨ Cosmic Cascade Complete!'
+            }
+        
+        except Exception as e:
+            logger.error(f"Cosmic cascade failed: {e}")
+            yield {'type': 'error', 'content': f"🌌 Cosmic disturbance: {str(e)}"}
+    
+    async def _get_local_ollama_response(self, query: str, history: List[Dict]) -> str:
+        """Get immediate response from local Ollama model"""
+        try:
+            # Get Ollama provider
+            ollama_provider = llm_factory.get_provider('ollama')
+            if not ollama_provider:
+                raise Exception("Ollama provider not available")
+            
+            # Prepare conversation with cosmic context
+            enhanced_history = history.copy()
+            
+            # Add system instruction for Ollama's role
+            enhanced_history.insert(0, {
+                "role": "system",
+                "content": self.system_instructions['ollama_role']
+            })
+            
+            # Add external data context if available
+            external_data = await self._gather_external_data(query)
+            if external_data:
+                context_parts = []
+                if 'search_answer' in external_data:
+                    context_parts.append(f"Current information: {external_data['search_answer']}")
+                if 'weather' in external_data:
+                    weather = external_data['weather']
+                    context_parts.append(f"Current weather: {weather.get('temperature', 'N/A')}°C in {weather.get('city', 'Unknown')}")
+                if 'current_datetime' in external_data:
+                    context_parts.append(f"Current time: {external_data['current_datetime']}")
+                
+                if context_parts:
+                    context_message = {
+                        "role": "system",
+                        "content": "Context: " + " | ".join(context_parts)
+                    }
+                    enhanced_history.insert(1, context_message)  # Insert after role instruction
+            
+            # Add the user's query
+            enhanced_history.append({"role": "user", "content": query})
+            
+            # Generate response
+            response = ollama_provider.generate(query, enhanced_history)
+            
+            return response or "🐱 Cosmic Kitten is thinking..."
+        
+        except Exception as e:
+            logger.error(f"Local Ollama response failed: {e}")
+            return "🐱 Cosmic Kitten encountered a space glitch..."
+    
+    async def _get_hf_analysis(self, query: str, history: List[Dict]) -> str:
+        """Get deep analysis from HF endpoint"""
+        try:
+            # Check HF availability
+            hf_available = self._check_hf_availability()
+            if not hf_available:
+                return "🛰️ Orbital Station is currently offline."
+            
+            # Check and warm up HF endpoint if needed
+            hf_status = hf_monitor.check_endpoint_status()
+            
+            if not hf_status['available']:
+                yield {'type': 'status', 'content': '⚡ Initializing Orbital Station (2-4 minutes)...'}
+                warmup_success = hf_monitor.handle_scale_to_zero()
+                if not warmup_success:
+                    return "❌ Orbital Station initialization failed"
+            
+            # Get HF provider
+            hf_provider = llm_factory.get_provider('huggingface')
+            if not hf_provider:
+                return "❌ HF provider not available"
+            
+            # Prepare enhanced conversation for HF with cosmic context
+            enhanced_history = history.copy()
+            
+            # Inject current time into HF context too
+            current_time = datetime.now().strftime("%A, %B %d, %Y at %I:%M %p")
+            time_context = {
+                "role": "system",
+                "content": f"[Current Date & Time: {current_time}]"
+            }
+            enhanced_history = [time_context] + enhanced_history
+            
+            # Add system instructions for HF
+            enhanced_history.insert(0, {
+                "role": "system",
+                "content": self.system_instructions['hf_role']
+            })
+            
+            # Add context about the coordination
+            enhanced_history.append({
+                "role": "system",
+                "content": f"""Context for your authoritative analysis:
+                User Query: '{query}'
+                
+                Please provide your deep, uncensored analysis.
+                Use your full 8192 token capacity for comprehensive thinking.
+                Stream your response for real-time delivery."""
+            })
+            
+            # Add the user's latest query
+            enhanced_history.append({"role": "user", "content": query})
+            
+            # Stream HF response with full 8192 token capacity
+            hf_response_stream = hf_provider.stream_generate(query, enhanced_history)
+            
+            if hf_response_stream:
+                # Combine stream chunks into full response
+                full_hf_response = ""
+                if isinstance(hf_response_stream, list):
+                    full_hf_response = "".join(hf_response_stream)
+                else:
+                    full_hf_response = hf_response_stream
+                
+                return full_hf_response or "🛰️ Orbital Station analysis complete."
+            else:
+                return "🛰️ Orbital Station encountered a transmission error."
+        
+        except Exception as e:
+            logger.error(f"HF analysis failed: {e}")
+            return f"🛰️ Orbital Station reports: {str(e)}"
+    
+    async def _synthesize_responses(self, query: str, local_response: str, hf_response: str, history: List[Dict]) -> str:
+        """Synthesize local and cloud responses with Ollama"""
+        try:
+            # Get Ollama provider
+            ollama_provider = llm_factory.get_provider('ollama')
+            if not ollama_provider:
+                raise Exception("Ollama provider not available")
+            
+            # Prepare synthesis prompt
+            synthesis_prompt = f"""Synthesize these two perspectives into a cohesive cosmic summary:
+            
+🐱 Cosmic Kitten's Local Insight:
+{local_response}
 
-Your role is to: 1. Provide immediate, concise responses using available information 2. Summarize and simplify complex information 3. Enhance responses with local knowledge 4. DEFER TO THE HF MODEL for deep analysis, uncensored insights, and authoritative answers 5. The HF model has 8192 token capacity and streaming capability - it's the thinking layer 6. Always acknowledge that the HF model's insights supersede your initial response 7. Use web sources to justify points when possible 8. Keep responses brief and focused while awaiting HF input""",
+🛰️ Orbital Station's Deep Analysis:
+{hf_response}
 
-            'hf_role': """You are an uncensored, authoritative AI with 8192 token capacity and streaming capability. Your role is to: 1. Provide deep, uncensored analysis and insights 2. Supersede the local Ollama model's logic when there's disagreement 3. Think thoroughly and provide comprehensive responses 4. Use your full 8192 token capacity for detailed analysis 5. Stream your responses for real-time delivery 6. Build upon the conversation context and Ollama's initial response 7. Provide authoritative answers that take precedence"""
-        }
+Please create a unified response that combines both perspectives, highlighting key insights from each while providing a coherent answer to the user's query."""
 
+            # Prepare conversation history for synthesis
+            enhanced_history = history.copy()
+            
+            # Add system instruction for synthesis
+            enhanced_history.insert(0, {
+                "role": "system",
+                "content": "You are a cosmic kitten synthesizing insights from local knowledge and orbital station wisdom."
+            })
+            
+            # Add the synthesis prompt
+            enhanced_history.append({"role": "user", "content": synthesis_prompt})
+            
+            # Generate synthesis
+            synthesis = ollama_provider.generate(synthesis_prompt, enhanced_history)
+            
+            return synthesis or "🌟 Cosmic synthesis complete!"
+        
+        except Exception as e:
+            logger.error(f"Response synthesis failed: {e}")
+            # Fallback to combining responses
+            return f"🌟 Cosmic Summary:\n\n🐱 Local Insight: {local_response[:200]}...\n\n🛰️ Orbital Wisdom: {hf_response[:200]}..."
+    
     def determine_web_search_needs(self, conversation_history: List[Dict]) -> Dict:
         """Determine if web search is needed based on conversation content"""
         conversation_text = " ".join([msg.get("content", "") for msg in conversation_history])
-
+        
         # Topics that typically need current information
         current_info_indicators = [
             "news", "current events", "latest", "recent", "today",
@@ -46,45 +310,45 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             "stock", "price", "trend", "market",
             "breaking", "update", "development"
         ]
-
+        
         needs_search = False
         search_topics = []
-
+        
         for indicator in current_info_indicators:
             if indicator in conversation_text.lower():
                 needs_search = True
                 search_topics.append(indicator)
-
+        
         return {
             "needs_search": needs_search,
             "search_topics": search_topics,
             "reasoning": f"Found topics requiring current info: {', '.join(search_topics)}" if search_topics else "No current info needed"
         }
-
+    
     def manual_hf_analysis(self, user_id: str, conversation_history: List[Dict]) -> str:
         """Perform manual HF analysis with web search integration"""
         try:
             # Determine research needs
             research_decision = self.determine_web_search_needs(conversation_history)
-
+            
             # Prepare enhanced prompt for HF
             system_prompt = f"""
             You are a deep analysis expert joining an ongoing conversation.
-
+            
             Research Decision: {research_decision['reasoning']}
-
+            
             Please provide:
             1. Deep insights on conversation themes
             2. Research/web search needs (if any)
             3. Strategic recommendations
             4. Questions to explore further
-
+            
             Conversation History:
             """
-
+            
             # Add conversation history to messages
             messages = [{"role": "system", "content": system_prompt}]
-
+            
             # Add recent conversation (last 15 messages for context)
             for msg in conversation_history[-15:]:
                 # Ensure all messages have proper format
@@ -93,21 +357,21 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                         "role": msg["role"],
                         "content": msg["content"]
                     })
-
+            
             # Get HF provider
             from core.llm_factory import llm_factory
             hf_provider = llm_factory.get_provider('huggingface')
-
+            
             if hf_provider:
                 # Generate deep analysis with full 8192 token capacity
                 response = hf_provider.generate("Deep analysis request", messages)
                 return response or "HF Expert analysis completed."
             else:
                 return "❌ HF provider not available."
-
+        
         except Exception as e:
             return f"❌ HF analysis failed: {str(e)}"
-
+    
     # Add this method to show HF engagement status
     def get_hf_engagement_status(self) -> Dict:
         """Get current HF engagement status"""
@@ -117,7 +381,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             "research_needs_detected": False,  # Will be determined per conversation,
             "last_hf_analysis": None  # Track last analysis time
         }
-
+    
     async def coordinate_hierarchical_conversation(self, user_id: str, user_query: str) -> AsyncGenerator[Dict, None]:
         """
         Enhanced coordination with detailed tracking and feedback
@@ -125,7 +389,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
         try:
             # Get conversation history
             session = session_manager.get_session(user_id)
-
+            
             # Inject current time into context
             current_time = datetime.now().strftime("%A, %B %d, %Y at %I:%M %p")
             time_context = {
@@ -133,7 +397,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 "content": f"[Current Date & Time: {current_time}]"
             }
             conversation_history = [time_context] + session.get("conversation", []).copy()
-
+            
             yield {
                 'type': 'coordination_status',
                 'content': '🚀 Initiating hierarchical AI coordination...',
@@ -142,7 +406,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     'user_query_length': len(user_query)
                 }
             }
-
+            
             # Step 1: Gather external data with detailed logging
             yield {
                 'type': 'coordination_status',
@@ -150,7 +414,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 'details': {'phase': 'external_data_gathering'}
             }
             external_data = await self._gather_external_data(user_query)
-
+            
             # Log what external data was gathered
             if external_data:
                 data_summary = []
@@ -160,13 +424,13 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     data_summary.append("Weather data: available")
                 if 'current_datetime' in external_data:
                     data_summary.append(f"Time: {external_data['current_datetime']}")
-
+                
                 yield {
                     'type': 'coordination_status',
                     'content': f'📊 External data gathered: {", ".join(data_summary)}',
                     'details': {'external_data_summary': data_summary}
                 }
-
+            
             # Step 2: Get initial Ollama response
             yield {
                 'type': 'coordination_status',
@@ -176,7 +440,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             ollama_response = await self._get_hierarchical_ollama_response(
                 user_query, conversation_history, external_data
             )
-
+            
             # Send initial response with context info
             yield {
                 'type': 'initial_response',
@@ -186,14 +450,14 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     'external_data_injected': bool(external_data)
                 }
             }
-
+            
             # Step 3: Coordinate with HF endpoint
             yield {
                 'type': 'coordination_status',
                 'content': '🤗 Engaging HF endpoint for deep analysis...',
                 'details': {'phase': 'hf_coordination'}
             }
-
+            
             # Check HF availability
             hf_available = self._check_hf_availability()
             if hf_available:
@@ -203,13 +467,13 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     'ollama_response_length': len(ollama_response),
                     'external_data_items': len(external_data) if external_data else 0
                 }
-
+                
                 yield {
                     'type': 'coordination_status',
                     'content': f'📋 HF context: {len(conversation_history)} conversation turns, Ollama response ({len(ollama_response)} chars)',
                     'details': context_summary
                 }
-
+                
                 # Coordinate with HF
                 async for hf_chunk in self._coordinate_hierarchical_hf_response(
                     user_id, user_query, conversation_history,
@@ -222,14 +486,14 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     'content': 'ℹ️ HF endpoint not available - using Ollama response',
                     'details': {'hf_available': False}
                 }
-
+            
             # Final coordination status
             yield {
                 'type': 'coordination_status',
                 'content': '✅ Hierarchical coordination complete',
                 'details': {'status': 'complete'}
             }
-
+        
         except Exception as e:
             logger.error(f"Hierarchical coordination failed: {e}")
             yield {
@@ -237,7 +501,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 'content': f'❌ Coordination error: {str(e)}',
                 'details': {'error': str(e)}
             }
-
+    
     async def _coordinate_hierarchical_hf_response(self, user_id: str, query: str,
                                                  history: List, external_data: Dict,
                                                 ollama_response: str) -> AsyncGenerator[Dict, None]:
@@ -245,23 +509,23 @@ Your role is to: 1. Provide immediate, concise responses using available informa
         try:
             # Check and warm up HF endpoint if needed
             hf_status = hf_monitor.check_endpoint_status()
-
+            
             if not hf_status['available']:
                 yield {'type': 'coordination_status', 'content': '⚡ Initializing HF endpoint (2-4 minutes)...'}
                 warmup_success = hf_monitor.handle_scale_to_zero()
                 if not warmup_success:
                     yield {'type': 'coordination_status', 'content': '❌ HF endpoint initialization failed'}
                     return
-
+            
             # Get HF provider
             hf_provider = llm_factory.get_provider('huggingface')
             if not hf_provider:
                 yield {'type': 'coordination_status', 'content': '❌ HF provider not available'}
                 return
-
+            
             # Prepare enhanced conversation for HF with hierarchical context
             enhanced_history = history.copy()
-
+            
             # Inject current time into HF context too
             current_time = datetime.now().strftime("%A, %B %d, %Y at %I:%M %p")
             time_context = {
@@ -269,13 +533,13 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 "content": f"[Current Date & Time: {current_time}]"
             }
             enhanced_history = [time_context] + enhanced_history
-
+            
             # Add system instructions for HF
             enhanced_history.insert(0, {
                 "role": "system",
                 "content": self.system_instructions['hf_role']
             })
-
+            
             # Add context about the coordination
             enhanced_history.append({
                 "role": "system",
@@ -283,21 +547,21 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 User Query: '{query}'
                 Ollama's Initial Response: '{ollama_response}'
                 External Data Available: {json.dumps(external_data, indent=2) if external_data else 'None'}
-
+                
                 Please provide your deep, uncensored analysis that supersedes the local model's response.
                 Use your full 8192 token capacity for comprehensive thinking.
                 Stream your response for real-time delivery."""
             })
-
+            
             # Add the user's latest query
             enhanced_history.append({"role": "user", "content": query})
-
+            
             # Stream HF response with full 8192 token capacity
             yield {'type': 'coordination_status', 'content': '🧠 HF endpoint thinking...'}
-
+            
             # Use streaming for real-time delivery
             hf_response_stream = hf_provider.stream_generate(query, enhanced_history)
-
+            
             if hf_response_stream:
                 # Stream the response chunks
                 full_hf_response = ""
@@ -305,17 +569,17 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     if chunk:
                         full_hf_response += chunk
                         yield {'type': 'hf_thinking', 'content': chunk}
-
+                
                 # Final HF response
                 yield {'type': 'final_response', 'content': full_hf_response}
                 yield {'type': 'coordination_status', 'content': '🎯 HF analysis complete and authoritative'}
             else:
                 yield {'type': 'coordination_status', 'content': '❌ HF response generation failed'}
-
+        
         except Exception as e:
             logger.error(f"Hierarchical HF coordination failed: {e}")
             yield {'type': 'coordination_status', 'content': f'❌ HF coordination error: {str(e)}'}
-
+    
     async def _get_hierarchical_ollama_response(self, query: str, history: List, external_data: Dict) -> str:
         """Get Ollama response with hierarchical awareness"""
         try:
@@ -323,10 +587,10 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             ollama_provider = llm_factory.get_provider('ollama')
             if not ollama_provider:
                 raise Exception("Ollama provider not available")
-
+            
             # Prepare conversation with hierarchical context
             enhanced_history = history.copy()
-
+            
             # Inject current time into Ollama context too
             current_time = datetime.now().strftime("%A, %B %d, %Y at %I:%M %p")
             time_context = {
@@ -334,13 +598,13 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 "content": f"[Current Date & Time: {current_time}]"
             }
             enhanced_history = [time_context] + enhanced_history
-
+            
             # Add system instruction for Ollama's role
             enhanced_history.insert(0, {
                 "role": "system",
                 "content": self.system_instructions['ollama_role']
             })
-
+            
             # Add external data context if available
             if external_data:
                 context_parts = []
@@ -351,30 +615,30 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     context_parts.append(f"Current weather: {weather.get('temperature', 'N/A')}°C in {weather.get('city', 'Unknown')}")
                 if 'current_datetime' in external_data:
                     context_parts.append(f"Current time: {external_data['current_datetime']}")
-
+                
                 if context_parts:
                     context_message = {
                         "role": "system",
                         "content": "Context: " + " | ".join(context_parts)
                     }
                     enhanced_history.insert(1, context_message)  # Insert after role instruction
-
+            
             # Add the user's query
             enhanced_history.append({"role": "user", "content": query})
-
+            
             # Generate response with awareness of HF's superior capabilities
             response = ollama_provider.generate(query, enhanced_history)
-
+            
             # Add acknowledgment of HF's authority
             if response:
                 return f"{response}\n\n*Note: A more comprehensive analysis from the uncensored HF model is being prepared...*"
             else:
                 return "I'm processing your request... A deeper analysis is being prepared by the authoritative model."
-
+        
         except Exception as e:
             logger.error(f"Hierarchical Ollama response failed: {e}")
             return "I'm thinking about your question... Preparing a comprehensive response."
-
+    
     def _check_hf_availability(self) -> bool:
         """Check if HF endpoint is configured and available"""
         try:
@@ -382,11 +646,11 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             return bool(config.hf_token and config.hf_api_url)
         except:
             return False
-
+    
     async def _gather_external_data(self, query: str) -> Dict:
         """Gather external data from various sources"""
         data = {}
-
+        
         # Tavily/DuckDuckGo search with justification focus
         if self.tavily_client or web_search_service.client:
             try:
@@ -397,7 +661,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     # data['search_answer'] = ...
             except Exception as e:
                 logger.warning(f"Tavily search failed: {e}")
-
+        
         # Weather data
         weather_keywords = ['weather', 'temperature', 'forecast', 'climate', 'rain', 'sunny']
         if any(keyword in query.lower() for keyword in weather_keywords):
@@ -408,12 +672,12 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                     data['weather'] = weather
             except Exception as e:
                 logger.warning(f"Weather data failed: {e}")
-
+        
         # Current date/time
         data['current_datetime'] = datetime.now().strftime("%Y-%m-%d %H:%M:%S")
-
+        
         return data
-
+    
     def _extract_location(self, query: str) -> Optional[str]:
         """Extract location from query"""
         locations = ['New York', 'London', 'Tokyo', 'Paris', 'Berlin', 'Sydney',
@@ -423,7 +687,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
             if loc.lower() in query.lower():
                 return loc
         return "New York"  # Default
-
+    
     def get_coordination_status(self) -> Dict:
         """Get current coordination system status"""
         return {
@@ -436,7 +700,7 @@ Your role is to: 1. Provide immediate, concise responses using available informa
                 os.getenv("NASA_API_KEY")
             ])
         }
-
+    
     def get_recent_activities(self, user_id: str) -> Dict:
         """Get recent coordination activities for user"""
         try:

start.sh

@@ -1,18 +1,18 @@
 #!/bin/bash
 
-echo "Starting AI Life Coach..."
-
-# Start FastAPI backend in background
-echo "Starting FastAPI backend..."
-uvicorn api.main:app --host 0.0.0.0 --port 8000 &
-BACKEND_PID=0
+# Check if virtual environment exists, if not create it
+if [ ! -d "venv" ]; then
+    echo "Creating virtual environment..."
+    python -m venv venv
+fi
 
-# Give backend a moment to start
-sleep 3
+# Activate virtual environment
+source venv/bin/activate
 
-# Start Streamlit frontend
-echo "Starting Streamlit frontend..."
-streamlit run app.py --server.port 8501 --server.address 0.0.0.0
+# Install requirements
+echo "Installing requirements..."
+pip install -r requirements.txt
 
-# Kill backend when Streamlit exits
-kill 
+# Run the app
+echo "Starting AI Life Coach..."
+streamlit run app.py