A detail that keeps getting buried high level: the agent is a wallet you can talk to. So given that assumption, and ignoring the gremlins about implementation of an embedded wallet vs BYOW. For an embedded wallet use case a user needs to:
- See a clear explanation of the agent wallet infrastructure ( e.g., why does
the agent have a wallet out of the box, what chains, why should i spin one up?
- Be guided to fund the agent's smart wallet for future transactions (see the
addresses and clear explanation of costs)
- also can we just create a wallet when someone creates their agent? pop the
interaction into a docker container once it transacts etc.