問題文
The org wants the agent kept within acceptable behavior across all its topics. Which combination of feature-level controls and platform protection does this?
選択肢
- Nothing can constrain the agent once enabled, and it says and does whatever it likes across all of the topics.
- The number of dashboards in the org, and a larger tile total is what keeps an agent within acceptable behavior, so topic scope and platform protection matter less.
- The course image assigned to the agent, and the artwork is what keeps its behavior within bounds.
- Topic scope and instructions constrain what the agent does and says, together with the Trust Layer's platform protections for data in the generative flow.