- Add GET /api/reports/last-sync endpoint to query server for last sync information
- Update reports page with sorting by client name and cap balance
- Create comprehensive client-side implementation guide (CLIENT_SYNC_IMPLEMENTATION_GUIDE.md)
- Create minimal client prompt (CLIENT_SYNC_MINIMAL_PROMPT.md)
- Update final implementation documentation (REPORTS_FINAL_IMPLEMENTATION.md)

This allows clients to:
- Query server for last sync information before syncing
- Verify local tracking against server state
- Recover from tracking corruption
- Prevent data loss from missed syncs

- Implement /api/reports/sync endpoint for client report data synchronization
- Add database models: ReportSync, Bet, BetDetail, ExtractionStats, ReportSyncLog
- Create web interface at /reports with filtering, pagination, and export (PDF/XLSX/CSV)
- Add sync logs interface at /sync-logs with comprehensive search and filtering
- Implement comprehensive logging for all sync operations
- Add migration 010 for existing systems to create all reports tables
- Update navigation menu with Reports and Sync Logs links
- Add weasyprint dependency for PDF generation
- Include test suite for reports sync API
- Add complete documentation in REPORTS_FEATURE_README.md

- Clients list page now removes ClientActivity records associated with inactive or revoked API tokens
- Only clients with active tokens are displayed in the web interface
- Added logging for cleanup operations
- Maintains existing filtering for active tokens in display

- Log request method, User-Agent header, and all headers on request arrival
- Helps debug what user agent the client is actually sending

- Log API token, rustdesk_id, IP, and user agent received
- Log client creation and updates with user agent changes
- Log database commit success
- Add rollback on exception for better error handling

- When rustdesk_id is not provided, update the most recent client for the API token
- This ensures user agents are updated even for clients not sending rustdesk_id
- Improves client reconnection handling

- Fix rustdesk_id retrieval for GET requests in /updates endpoint
- Ensure user agent is properly updated when clients reconnect
- Previously, GET requests defaulted to 'unknown' rustdesk_id, preventing proper client identification

- Added CSRF exemption to all JWT-authenticated API endpoints
- API routes use JWT tokens, not session cookies, so CSRF protection is not needed

- Change import to from flask_wtf.csrf import csrf_exempt
- Replace @csrf.exempt with @csrf_exempt decorators

- Initialize CSRFProtect in app/__init__.py
- Update import in app/upload/routes.py from flask_wtf.csrf to flask_wtf
- Maintain existing @csrf.exempt decorators

- Improved FileLike class with proper file handling and close() method
- Added better error handling and validation for JSON input
- Added try/finally block to ensure file handles are closed
- Added detailed error logging with traceback for debugging
- Ensures server always returns JSON instead of HTML error pages

- Error message now shows correct maximum file size from database setting
- Consistent with other validation functions using the same setting

- validate_file_size() now uses database setting instead of MAX_CONTENT_LENGTH
- detect_malicious_content() uses database setting for size limits
- Both functions now respect the configurable max_upload_size_mb (default 2048MB = 2GB)

- Explicitly set MAX_CONTENT_LENGTH in Flask app config during initialization
- Add logging to verify MAX_CONTENT_LENGTH is set correctly
- Fixes 'File too large' error for files over 500MB (now supports 5GB)

- Add seek(), tell(), read() methods to FileLike class
- Add content_type attribute for validation
- Fixes 'FileLike object has no attribute seek' error in chunked uploads