For the moment I consider that only one agent can be active. If multiple are detected, I raise an error. That is intentional to gather more context and feedback

For the moment I consider that only one agent can be active. If multiple are detected, I raise an error. That is intentional to gather more context and feedback from users. I will change this behavior as soon as someone report me a use-case. See the implementation at: https://github.com/kdeldycke/extra-platforms/blob/ecbe740bf7abe8b027ddda8512f3019f2e5bb02a/extra_platforms/detection.py#L1609-L1651 https://github.com/kdeldycke/extra-platforms/blob/ecbe740bf7...