pan-relay

Pan-Relay is a multi-drive aggregation relay and automated monetization management system built with Python and Flask, primarily aimed at cloud drive promoters and resource site owners. Its core workflow is "resource aggregation -> automatic re-saving -> monetization link replacement -> priority distribution", re-saving external drive resources to a personal drive and then generating one's own share links for user acquisition and re-save revenue. The project has been adapted for Quark Drive, Baidu Netdisk, Aliyun Drive, UC Drive and Xunlei Drive, supporting automatic identification, re-saving and temporary sharing. The system has a built-in SQLite database, requiring no additional configuration of MySQL or other database services, and automatically initializes table structures and preset search sources on startup. For search, the frontend first prioritizes links from the private monetization resource library, then concurrently aggregates third-party APIs, public Telegram channels and extensible Python search plugins. Telegram channel search requires no Bot Token and can automatically extract drive links and extraction codes. External search results support on-demand re-saving, with fallback to the original link on failure, and expired shares are handled by a scheduled cleanup service. The project provides an admin backend for resource management, search API configuration, drive credential configuration and frontend drive display control. The default backend account is admin with password admin123. It also provides a standardized /api/v1 REST API using a two-step process: first aggregate query, then automatic re-save and replacement, making it easy to integrate with WeChat mini-programs, mobile apps, Telegram bots or resource navigation sites. The frontend and backend interfaces can be independently enabled or disabled via environment variables, supporting headless mode operation. Deployment options include one-click startup with Docker Compose and running in a local Python environment, with the service listening on port 5004 by default. The project is open-sourced under the MIT license and states that it is for technical exchange and learning research only.