ForkForensics

ForkForensics is a native desktop application designed to monitor and recover GitHub repositories that may have vanished unexpectedly. As repositories can silently disappear—a result of deletion or changes in visibility—ForkForensics tracks these changes, working diligently to reclaim historical data before it is permanently lost.

Key Features

• Automated Daily Monitoring: ForkForensics automatically ingests data from the GH Archive every night, verifying up to 5,000 repositories per cycle without requiring manual polls.
• Dual Recovery Signals: The application employs two independent recovery techniques, ranking forks by their commit history depth while probing for orphaned commits still accessible by SHA.
• Custom Watchlist: Users can compile a watchlist of priority repositories, ensuring they are always included in the verification queue regardless of their activity status on GitHub.
• Local Recovery: All cloning and recovery processes occur locally; no data is pushed to GitHub, maintaining user privacy and control.
• Multi-Token Rotation: The app supports configuring multiple GitHub tokens for seamless operation, automatically rotating tokens to prevent hitting usage limits.
• Desktop Notifications: Users receive notifications in the system tray about the results of the nightly monitoring cycle.

Operational Workflow

ForkForensics runs a scheduled check each night, downloading the previous day's GitHub Archive data. It then re-verifies repositories to identify disappearances and automatically begins recovery investigations for repositories no longer accessible. The recovery process involves:

1. Identifying forks of the repository and ranking them by historical commit depth—deeper forks are prioritized as the best candidates for recovery.
2. Investigating orphan commits through independent signals to locate data that may remain retrievable, compiling a detailed report of findings for each repository.

Background

This project was conceived from the need to restore critical historical data for a quantitative trading initiative after a relied-upon repository was compromised. ForkForensics streamlines this manual process into an automated, user-friendly experience. An initial test validated that the app's algorithm could reproduce a manually conducted recovery ranking, demonstrating its reliability.

Important Considerations

• Scale Limitations: The volume of data in the GH Archive can sometimes exceed the capabilities of a single GitHub token within a day, necessitating the use of multiple tokens to expand capacity.
• Indistinguishable States: The application cannot differentiate between deleted and private repositories, both of which appear as "vanished".
• Orphan Commit Duration: The retention period for orphan commits by GitHub is not guaranteed, so the app provides current data without long-term assurances.

Security

Tokens used within ForkForensics are stored unencrypted locally and should be handled with care. The application performs all operations without pushing data to external repositories, ensuring that the monitored data remains private and secure.

For further details on the project, including installation and testing instructions, please refer to the ForkForensics GitHub repository.