Design a capability discovery model so agents can use new tools without becoming unstable.

Instruction: Describe how you would add new tools while keeping model behavior predictable.

Context: Assesses whether the candidate can design a practical architecture and explain the main tradeoffs. Describe how you would add new tools while keeping model behavior predictable.

Official answer available

Preview the opening of the answer, then unlock the full walkthrough.

I would separate discovery from eligibility. Agents can learn that a capability exists, but they should only consider it usable when the host has approved it for the current workflow, context, and risk level.

I also want capability metadata that tells the model what the tool is for,...

Upgrade to view official answer

Related Questions