feat: initial IQAI multi-model AI dashboard

- Express backend with Replicate API proxy (chat, models, account, search)
- React + Vite + Tailwind frontend with custom Midnight Violet color scheme
- @mention autocomplete to route messages to specific models
- Parallel multi-model queries with model selection in sidebar
- DuckDuckGo web search context injection
- Model manager UI (add/edit/remove Replicate models)
- Per-model system instructions per conversation
- Replicate account info display in sidebar
- Conversation history with local persistence (Zustand)
- Full Docker deployment (backend + nginx-served frontend)
- Montserrat + Poppins fonts

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

backend/routes/account.js

@@ -0,0 +1,35 @@
+import express from 'express';
+
+const router = express.Router();
+const REPLICATE_BASE = 'https://api.replicate.com/v1';
+
+router.get('/', async (req, res) => {
+  const token = process.env.REPLICATE_API_TOKEN;
+  if (!token) return res.status(400).json({ error: 'REPLICATE_API_TOKEN not configured' });
+
+  try {
+    const response = await fetch(`${REPLICATE_BASE}/account`, {
+      headers: { 'Authorization': `Bearer ${token}` }
+    });
+    if (!response.ok) {
+      const err = await response.json().catch(() => ({}));
+      return res.status(response.status).json({ error: err.detail || 'Failed to fetch account' });
+    }
+    const account = await response.json();
+
+    // Try to get hardware/usage info
+    let hardware = null;
+    try {
+      const hwRes = await fetch(`${REPLICATE_BASE}/hardware`, {
+        headers: { 'Authorization': `Bearer ${token}` }
+      });
+      if (hwRes.ok) hardware = await hwRes.json();
+    } catch (_) {}
+
+    res.json({ account, hardware });
+  } catch (err) {
+    res.status(500).json({ error: err.message });
+  }
+});
+
+export { router as accountRouter };