perf: Phase 1+2+3 — Turbopack, Prisma select, RSC page, CSS masonry + dnd-kit

- Turbopack activé (dev: next dev --turbopack)
- NOTE_LIST_SELECT: exclut embedding (~6KB/note) des requêtes de liste
- getAllNotes/getNotes/getArchivedNotes/getNotesWithReminders optimisés
- searchNotes: filtrage DB-side au lieu de full-scan JS en mémoire
- getAllNotes: requêtes ownNotes + sharedNotes parallélisées avec Promise.all
- syncLabels: upsert en transaction () vs N boucles séquentielles
- app/(main)/page.tsx converti en Server Component (RSC)
- HomeClient: composant client hydraté avec données pré-chargées
- NoteEditor/BatchOrganizationDialog/AutoLabelSuggestionDialog: lazy-loaded avec dynamic()
- MasonryGrid: remplace Muuri par CSS grid auto-fill + @dnd-kit/sortable
- 13 packages supprimés: muuri, web-animations-js, react-masonry-css, react-grid-layout
- next.config.ts nettoyé: suppression webpack override, activation image optimization

mcp-server/package.json

@@ -1,13 +1,16 @@
 {
   "name": "memento-mcp-server",
-  "version": "3.0.0",
-  "description": "MCP Server for Memento - AI-powered note-taking app. Provides 34 tools for notes, notebooks, labels, AI features, and reminders.",
+  "version": "3.1.0",
+  "description": "MCP Server for Memento - AI-powered note-taking app. Optimized with connection pooling, batch operations, and caching. Provides 37 tools for notes, notebooks, labels, AI features, and reminders.",
   "type": "module",
   "main": "index.js",
   "scripts": {
     "start": "node index.js",
     "start:http": "node index-sse.js",
-    "start:sse": "node index-sse.js"
+    "start:sse": "node index-sse.js",
+    "dev": "MCP_LOG_LEVEL=debug node index-sse.js",
+    "test:perf": "node test/performance-test.js",
+    "test:connection": "node test/connection-test.js"
   },
   "dependencies": {
     "@modelcontextprotocol/sdk": "^1.0.4",
@@ -19,5 +22,16 @@
   "devDependencies": {
     "@types/node": "^20.0.0",
     "prisma": "^5.22.0"
+  },
+  "keywords": [
+    "mcp",
+    "memento",
+    "notes",
+    "ai",
+    "optimized",
+    "performance"
+  ],
+  "engines": {
+    "node": ">=18.0.0"
   }
 }