It's an interesting addition to their security model. However I don't like how a number of their examples use useragent as a 'secure' way to provide secured el

It's an interesting addition to their security model. However I don't like how a number of their examples use useragent as a 'secure' way to provide secured elevated privileges on an object via a specific app or location. user agents are fungible and should not be used for security permissions. I'm surprised Amazon used these in this way in their examples.