Phase 10: Per-User Rate Limits — messages + tokens, quota UI, admin usage

Backend:
- max_ai_messages_per_day + max_ai_tokens_per_day on User model (nullable, override)
- Migration 008: add columns + seed default settings (100 msgs, 500K tokens)
- usage_service: count today's messages + tokens, check quota, get limits
- GET /chats/quota returns usage vs limits + reset time
- POST /chats/{id}/messages checks quota before streaming (429 if exceeded)
- Admin user schemas expose both limit fields
- GET /admin/usage returns per-user daily message + token counts
- admin_user_service allows updating both limit fields

Frontend:
- Chat header shows "X/Y messages · XK/YK tokens" with red highlight at limit
- Quota refreshes every 30s via TanStack Query
- Admin usage page with table: user, messages today, tokens today
- Route + sidebar entry for admin usage
- English + Russian translations

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

backend/app/models/user.py

@@ -16,6 +16,8 @@ class User(Base):
     role: Mapped[str] = mapped_column(String(20), nullable=False, default="user")
     is_active: Mapped[bool] = mapped_column(Boolean, nullable=False, default=True)
     max_chats: Mapped[int] = mapped_column(Integer, nullable=False, default=10)
+    max_ai_messages_per_day: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    max_ai_tokens_per_day: Mapped[int | None] = mapped_column(Integer, nullable=True)
     oauth_provider: Mapped[str | None] = mapped_column(String(50), nullable=True)
     oauth_provider_id: Mapped[str | None] = mapped_column(String(255), nullable=True)
     telegram_chat_id: Mapped[int | None] = mapped_column(nullable=True)