关于
This skill focuses on transforming experimental Python scripts into professional, reproducible research code. It automates the centralization of runtime parameters into structured configuration objects, implements rigorous seeding across Python and framework-specific random number generators, and establishes a systematic workflow for recording environment metadata like Git hashes and configuration snapshots. By standardizing output layouts and removing hidden global states, it ensures that every experiment run is traceable, auditable, and perfectly repeatable across different environments.
