feat(whatsapp): QWEN primary AI backend, Hermes fallback, conversation history, vehicle persistence, demo prompts

- Add QWEN (qwen3.6) as primary AI backend with short system prompt
- Hermes remains as fallback with 45s timeout
- Increase QWEN timeout to 35s, max_tokens to 4000
- Add conversation history loading from whatsapp_messages (last 4 msgs)
- Persist detected vehicle in whatsapp_sessions table
- Add 'limpiar chat' / 'nuevo chat' / 'reset' commands to clear history
- Fix CSS conflict: rename whatsapp chat-panel classes to wa-chat-panel
- Fix JS ID conflicts with chat.js widget (waChatPanel, waChatMessages, etc.)
- Improve no-stock response: conversational with alternatives
- Split search_query by | for multi-part lookups
- Add DEMO_PROMPTS.md and DEMO_PROMPTS_V2.md

pos/static/pwa/sw.js

@@ -2,7 +2,7 @@
 // Nexus POS — Service Worker v3
 // Self-contained vanilla JS. No external imports.
 
-const CACHE_NAME = 'nexus-pos-v3';
+const CACHE_NAME = 'nexus-pos-v4';
 
 const APP_SHELL = [
   '/pos/login',
@@ -176,7 +176,11 @@ self.addEventListener('fetch', function (event) {
     return;
   }
 
-  // API calls → network-first
+  // API calls → network-first (except WhatsApp which must be real-time)
+  if (url.pathname.indexOf('/pos/api/whatsapp/') !== -1) {
+    // WhatsApp endpoints need fresh server data; skip SW caching
+    return;
+  }
   if (url.pathname.indexOf('/pos/api/') !== -1) {
     event.respondWith(networkFirst(req));
     return;