
@adrishaBiswas We make decisions using the agent’s live context, not just predefined rules. Things like user permissions, prior actions, conversation state, and external data can all influence whether an action is allowed, blocked, or escalated.
English
























