Ironic, but on the rails, I guess.
Many more improvements are needed in terms of agent use cases. It's good to work with them, but they need extra testing before fully trusting them; otherwise, like in this case, even if the findings help, it can be tedious trying to distinguish what is real and what is not.