Create ui/lifeadmin_coach_ui.py

ui/lifeadmin_coach_ui.py

@@ -0,0 +1,224 @@
+"""
+LifeAdmin Coach UI - AI Chatbot Assistant
+Provides conversational interface with RAG context and memory
+"""
+
+import gradio as gr
+import asyncio
+from datetime import datetime
+
+
+def create_lifeadmin_coach_ui(agent):
+    """Create LifeAdmin Coach chatbot interface"""
+    
+    async def chat_with_coach(message, history):
+        """Process chat message with RAG and memory context"""
+        if not message or not message.strip():
+            return history, ""
+        
+        try:
+            # Get relevant context from RAG
+            rag_context = ""
+            try:
+                rag_results = await agent.rag_engine.search(message, k=3)
+                if rag_results:
+                    rag_context = "\n".join([
+                        f"- {doc.get('text', '')[:200]}..." 
+                        for doc in rag_results
+                    ])
+            except Exception:
+                pass
+            
+            # Get relevant memories
+            memory_context = ""
+            try:
+                memory_context = agent.memory.get_relevant_memories(message, k=3)
+            except Exception:
+                pass
+            
+            # Build enhanced prompt
+            system_context = f"""You are LifeAdmin Coach, a helpful AI assistant that helps users manage their life admin tasks.
+
+You have access to:
+- User's uploaded documents and files (via RAG)
+- Past conversation history (via memory)
+- Various tools for task automation
+
+Your role:
+- Answer questions about uploaded documents
+- Provide advice on life admin tasks (bills, forms, scheduling, etc.)
+- Help organize and plan tasks
+- Be friendly, concise, and actionable
+
+"""
+            
+            if rag_context:
+                system_context += f"\n📚 **Relevant Documents:**\n{rag_context}\n"
+            
+            if memory_context:
+                system_context += f"\n💭 **Past Context:**\n{memory_context}\n"
+            
+            # Build full prompt
+            full_prompt = f"""{system_context}
+
+**User Question:** {message}
+
+Provide a helpful, concise response. If the user asks about their documents, reference the relevant documents context above.
+"""
+            
+            # Get LLM response
+            from utils.llm_utils import get_llm_response
+            response = await get_llm_response(full_prompt, temperature=0.7, max_tokens=1000)
+            
+            # Save to memory
+            try:
+                agent.memory.add_memory(
+                    content=f"User: {message}\nCoach: {response}",
+                    memory_type='conversation',
+                    metadata={'timestamp': datetime.now().isoformat()},
+                    importance=5
+                )
+            except Exception:
+                pass
+            
+            # Update history
+            history.append({"role": "user", "content": message})
+            history.append({"role": "assistant", "content": response})
+            
+            return history, ""
+            
+        except Exception as e:
+            error_msg = f"⚠️ Error: {str(e)}"
+            history.append({"role": "user", "content": message})
+            history.append({"role": "assistant", "content": error_msg})
+            return history, ""
+    
+    async def quick_action(action_type, history):
+        """Handle quick action buttons"""
+        actions = {
+            "summarize_docs": "Can you summarize all the documents I've uploaded?",
+            "list_tasks": "What tasks or action items can you find in my documents?",
+            "deadlines": "Are there any deadlines or important dates I should know about?",
+            "help": "What can you help me with?"
+        }
+        
+        message = actions.get(action_type, "")
+        if message:
+            return await chat_with_coach(message, history)
+        return history, ""
+    
+    def clear_chat():
+        """Clear chat history"""
+        return [], ""
+    
+    # Build UI
+    with gr.Column():
+        gr.Markdown("""
+        ## 🤖 LifeAdmin Coach
+        
+        Your AI assistant for managing life admin tasks. Ask questions about your documents,
+        get advice on tasks, or chat about anything related to organizing your life!
+        
+        **I can help you with:**
+        - Answering questions about uploaded documents
+        - Finding deadlines and important dates
+        - Organizing tasks and creating plans
+        - Drafting emails and messages
+        - General life admin advice
+        """)
+        
+        # Chat interface
+        chatbot = gr.Chatbot(
+            label="💬 Chat with LifeAdmin Coach",
+            height=500,
+            type="messages"
+        )
+        
+        # Input area
+        with gr.Row():
+            msg_input = gr.Textbox(
+                label="",
+                placeholder="Ask me anything about your documents or life admin tasks...",
+                scale=4,
+                lines=2
+            )
+            send_btn = gr.Button("Send", variant="primary", scale=1)
+        
+        # Quick action buttons
+        with gr.Row():
+            gr.Markdown("**Quick Actions:**")
+        
+        with gr.Row():
+            summarize_btn = gr.Button("📄 Summarize Docs", size="sm")
+            tasks_btn = gr.Button("✅ List Tasks", size="sm")
+            deadlines_btn = gr.Button("📅 Find Deadlines", size="sm")
+            help_btn = gr.Button("❓ Help", size="sm")
+            clear_btn = gr.Button("🗑️ Clear Chat", size="sm", variant="stop")
+        
+        # Examples
+        with gr.Accordion("💡 Example Questions", open=False):
+            gr.Examples(
+                examples=[
+                    "What documents have I uploaded?",
+                    "Are there any bills I need to pay?",
+                    "Help me organize my tasks for this week",
+                    "Draft an email to follow up on my membership application",
+                    "What are the key points from my PDF?",
+                    "Find all phone numbers and emails in my documents",
+                    "When is my next deadline?",
+                ],
+                inputs=msg_input
+            )
+        
+        # Event handlers
+        def chat_wrapper(message, history):
+            """Sync wrapper for async chat function"""
+            return asyncio.run(chat_with_coach(message, history))
+        
+        def quick_action_wrapper(action_type, history):
+            """Sync wrapper for quick actions"""
+            return asyncio.run(quick_action(action_type, history))
+        
+        # Connect events
+        msg_input.submit(
+            fn=chat_wrapper,
+            inputs=[msg_input, chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        send_btn.click(
+            fn=chat_wrapper,
+            inputs=[msg_input, chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        summarize_btn.click(
+            fn=lambda h: quick_action_wrapper("summarize_docs", h),
+            inputs=[chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        tasks_btn.click(
+            fn=lambda h: quick_action_wrapper("list_tasks", h),
+            inputs=[chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        deadlines_btn.click(
+            fn=lambda h: quick_action_wrapper("deadlines", h),
+            inputs=[chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        help_btn.click(
+            fn=lambda h: quick_action_wrapper("help", h),
+            inputs=[chatbot],
+            outputs=[chatbot, msg_input]
+        )
+        
+        clear_btn.click(
+            fn=clear_chat,
+            outputs=[chatbot, msg_input]
+        )
+    
+    return gr.Column()