This is very cool. I'm not sure what the path is from "import this github library into your code, listen for phrase 'restart hermes nodes 1 through 4' and trigger appropriate actions" but this is still exciting to see this on-device voice recognition ported to so many platforms and with basic documentation for each.