Lastly, you can now give a Claude Managed Agent an advisor: a stronger model it can call mid-session for a second opinion. Add one line to the roster and the working agent gets an advisor.
Anthropic added an advisor option to Claude Managed Agents: the working AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → can consult a stronger model during a session. The announcement's example configures an API-design agent to seek advice before irreversible decisions. This provides a way to keep one model responsible for the ongoing task while making another available for difficult judgments. The supported capability is consultation on demand; the announcement does not report an improvement in task success or a measured cost saving.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
