You start with this premise
It’s great and works really well for us, but it raises lots of questions. Does everyone get a bot? If people share one, whose credentials does it use?
Eventually you throw in this admission in the middle of other stuff:
Buzz gives each agent its own key. The agent’s owner signs a narrowly scoped authorization. The agent then signs its own work with its own identity.
So you're explicitly picking a bot per person, you're just calling them an agent. But you never really admit this because you call it 'an agent' rather than "the user's agent". You're still ending up with lots of agents each owned by people, you're just writing another orchestration layer yourself.
The agent remains the author. Its credential proves who authorized it and under what conditions.
Right yeh, so the human is the responsible party.
If an agent key leaks, revoke the agent without replacing the human identity behind it.
Just like you revoke an app-key on GitHub/Slack. This isn't a revolutionary thing you're doing.
You then throw in this curveball.
Buzz can also run an agent’s model requests on another community member’s machine.
How? How is this resource managed or limited? Do I have to assume Buzz will just borrow my laptop's CPU & GPU at random?