[Evaluation] Add OSWorld Benchmark

[xingyaoww]

What problem or use case are you trying to solve?

Adding OSWorld benchmark to OpenHands evaluation harness evaluation/benchmarks.

Describe the UX of the solution you'd like

Do you have thoughts on the technical implementation?

The primary challenge is to be able to emulate an OS inside the Docker-based runtime. Fortunately, OSWorld authors already figured out a way to do it by running qemu inside docker:

https://github.com/xlang-ai/OSWorld?tab=readme-ov-file#docker-server-with-kvm-support-for-the-better

This likely needs major work with our runtime to support -- but once they are working, OpenHands runtime will be much more capable since we can optionally have access to a full OS (with GUI) for both agents and humans to interact with.

Describe alternatives you've considered

Additional context