Tag
An engineering discussion exploring how to change an AI agent's authority and permissions after it's already deployed in production, comparing approaches like config deploys, IAM/RBAC, policy engines such as Cedar and OPA, and gateway/proxy layers.