Eval

[polina-c]

Contributes to #764

Docs: https://docs.github.com/en/actions/how-tos/write-workflows/choose-what-workflows-do/use-secrets

[gemini-code-assist]

Code Review

I have read and understood ./docs/contributing/README.md.

This pull request introduces a new eval example project for evaluating genui against AI models. The setup is straightforward, but I've identified a few areas for improvement. My main feedback concerns code duplication for API key retrieval, which is also present in other examples. Addressing this would improve maintainability. I've also pointed out an unused asset configuration and a minor naming issue. Overall, a good start for the new evaluation project.