implement backend skeleton with MongoDB and Claude AI integration

- Add controllers (Auth, Chat, Event) with placeholder implementations
- Add services (Auth, Chat, Event) with business logic interfaces
- Add repositories with MongoDB/Mongoose models (User, Event, Chat)
- Add middleware for JWT authentication
- Add Claude AI adapter implementing AIProvider interface
- Add utility modules for JWT and password handling
- Add shared types and DTOs for User, CalendarEvent, ChatMessage
- Configure routes with proper endpoint structure
- Update app.ts with dependency injection setup
- Add required dependencies: mongoose, bcrypt, jsonwebtoken, @anthropic-ai/sdk

apps/server/src/repositories/mongo/MongoChatRepository.ts

@@ -0,0 +1,23 @@
+import { ChatMessage, Conversation, CreateMessageDTO, GetMessagesOptions } from '@caldav/shared';
+import { ChatRepository } from '../../services/interfaces';
+import { ChatMessageModel, ConversationModel } from './models';
+
+export class MongoChatRepository implements ChatRepository {
+  // Conversations
+  async getConversationsByUser(userId: string): Promise<Conversation[]> {
+    throw new Error('Not implemented');
+  }
+
+  async createConversation(userId: string): Promise<Conversation> {
+    throw new Error('Not implemented');
+  }
+
+  // Messages (cursor-based pagination)
+  async getMessages(conversationId: string, options?: GetMessagesOptions): Promise<ChatMessage[]> {
+    throw new Error('Not implemented');
+  }
+
+  async createMessage(conversationId: string, message: CreateMessageDTO): Promise<ChatMessage> {
+    throw new Error('Not implemented');
+  }
+}