Migrate to SSR with archive system, auth, and EC2 deployment

[jploskey]

Migrates the dashboard from a static SPA to a server-rendered Nuxt app running on EC2, and adds a full archive and authentication system.

• SSR enabled with Nitro server engine, Caddy reverse proxy, and systemd service
• SQLite database with auto-migrations for users, sessions, and archive records
• Cookie-based auth with scrypt password hashing, admin/viewer roles, login page, and user management UI
• S3 proxy middleware for plot images, discrete data, spectrograms, echograms, and archives
• Three archive types: scheduled (automatic daily via cron), event (user-created), and internal (admin-only, auth-gated)
• Archives stored in S3 under archives/<type>/<key>/QAQC_plots/ with automatic retention policy
• Archive viewer with sidebar dropdown, banner, type filter tabs persisted in URL, and conflict detection
• Archives page with create/delete dialogs, role-based visibility, and SSR-compatible data loading
• Centralized useAuth Pinia store with SSR state hydration via callOnce
• usePersisted composable reads URL query params during SSR to avoid hydration mismatches
• Trailing slash redirect middleware to prevent 404s from cached redirects
• Scheduled tasks for archive cleanup, database backup to S3, and session expiry
• Configurable archive schedule via QAQC_ARCHIVE_SCHEDULE environment variable
• CI/CD workflow builds on GitHub Actions and deploys built output to EC2 via SCP
• Documentation covering environment variables, S3 bucket structure, auth, server logs, and troubleshooting