Hello,
I'm using a LLM agent to help me test ideas with Zorro and I'm encountering an issue. When there are errors they only appear in the GUI terminal and are not accessible from the command line. For this reason the LLM cannot know what went wrong and needs me to look at the GUI and tell it. I'm trying to automate the boring things as much as possible and would like to find a solution.
Do you see a way to access the errors from the terminal, maybe some undocumented parameter? Thank you.