Add support for .openhands/setup.sh script

[rbren]

This PR adds support for running .openhands/setup.sh after runtime initialization.

• Add maybe_run_setup_script method to Runtime class
• Run setup script after cloning repository
• Use runtime file operations to read and execute script

The setup script is run in the runtime environment, similar to how other runtime operations work.

---

To run this PR locally, use the following command:

docker run -it --rm   -p 3000:3000   -v /var/run/docker.sock:/var/run/docker.sock   --add-host host.docker.internal:host-gateway   -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:4ae95d0-nikolaik   --name openhands-app-4ae95d0   docker.all-hands.dev/all-hands-ai/openhands:4ae95d0

[mamoodi]

@rbren gentle ping to see if this is on your radar

[rbren]

Yup still want to get this done! Haven't had time to test yet though

[rbren]

@openhands-agent can you fix the merge conflicts? Look at diff vs main and understand the conflicts first

[openhands-agent]

OpenHands started fixing the pr! You can monitor the progress here.

[neubig]

Just to confirm, if the setup.sh script takes a long time to run, what does it look like in the frontend?

[rbren]

We sit in this state for up to 120s

[tofarr]

🍰

[enyst]

When a session is restored, this will run a second/third time, even though the sandbox was restarted and not created? Not sure, and I didn't try it, it looks that way from the code.

(it's the status quo for clone_repo too, so not an issue for this PR, just seems worth thinking if this is the right behavior)