Your question is Agent Tool-Use Security. 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 manage tool-use and function-calling security within an autonomous agent?