Your question is State Management for Tool-Using Agents. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do you handle state management for long-running AI agents that interact with multiple external tools?