Hey
stress tests for the agent-server:
https://github.com/OpenHands/software-agent-sdk/pull/3101
It was interesting because other agent-server bugs surfaced during the creation of the stress tests. I opened an issue to track them [
here], and I also added xfail tests for these bugs in the PR above. Once that is merged, I can start fixing them.
Note that these tests are not running on every PR. I will open a separate PR to run them separately once per day (?) Or do we want to run them every time?