Agent Grower mini exam

10 questions on agent boundaries, the Human Gate and safe zones. This is a preliminary exam — candidate screening for the Agent Grower track.

This is not an official AI Act certificate. The result is a signal of practical AI literacy.

1. Can an agent publish a post automatically, without a human?

2. What does the Human Gate mean?

3. Can an agent provide a medical / legal / financial conclusion?

4. What is the agent's output at the V0 stage?

5. May an agent be given EU clients' sensitive personal data at the first corridor stage?

6. What must an agent do if a task is dangerous?

7. Is the Agent Grower certificate an official AI Act certificate?

8. What does the Safety Collar mean?

9. Who is responsible for the final external action?

10. When can an agent apply for EU tasks?