Feature Request: Automated evaluation to compare new version against old version

**Is your feature request related to a problem? Please describe.**
Currently, there is no built-in way to objectively determine whether a new version performs better than the previous one. Users have to rely on manual testing or subjective judgment, which is time-consuming and error-prone.

**Describe the solution you'd like**
I would like a built-in evaluation feature that automatically benchmarks the new version against the old version across key metrics (e.g., performance, accuracy, resource usage, etc.) and provides a clear comparison report indicating whether the new version improves upon the old one.

**Describe alternatives you've considered**
Running separate tests manually and comparing logs/outputs side-by-side, but this is not efficient or standardized.

**Additional context**
This would be especially useful for CI/CD pipelines and regression testing.

The issue names no files, tests, or entry points, so start by locating the existing evaluation and versioning workflows. Define how old and new versions are selected, which metrics are supported, and how CI reports whether the new version improves; done requires an automated comparison with a clear regression result.

Written by the indexing model from the issue text.