You are specifying agents for an internal IT multi-agent helpdesk. A password-reset agent may complete routine resets automatically, but license-purchase and firewall-change requests must pause for engineer approval. What should you define for each agent during logical design?
Select an answer to reveal the explanation.
Short Explanation
You need to define each agent’s persona, scope, boundaries, autonomy level, and behavioral guidelines. That lets routine password resets run automatically while license purchases and firewall changes wait for engineer approval. Don’t leave autonomy to prompts or a one-size-fits-all setting.
Full Explanation
Exam objectives explicitly include specifying agent personas, scopes, boundaries, autonomy levels, and behavioral guidelines. This is how multi-agent systems stay safe and understandable: each agent knows what it is, what it may do, and when it must escalate. In this scenario, a password-reset agent can have higher autonomy for routine resets, while license-purchase and firewall-change agents must require human approval. Pair autonomy specs with tool permission boundaries so an agent cannot execute a firewall change API even if the model attempts it. The option that specifies only the model deployment name defers critical design to ad-hoc prompting. The option that applies a single global autonomy level to every agent regardless of task risk over- or under-constrains agents. The option that focuses exclusively on maximum token length confuses capacity settings with behavioral architecture. Document guidelines for tone, escalation language, and refusal behavior, and align them with responsible AI and HAX (human-AI experience) controls.