AI agents practice assessment. Foundations

This is a scored practice assessment. It is not timed. Use it to sharpen your judgement before a formal assessment.

Answered 0 of 8

Pick one option per question. Your choices save automatically in this browser. You will see your score and recommendations after you submit.

Scenario. An agent keeps looping between tools and never answers. What is the first control to add?

Unansweredcontrol

Scenario. A tool returns content from an untrusted source. What is the safest default?

Unansweredsafety

Scenario. You need the agent to remember a user preference across sessions. Which memory type fits best?

Unansweredmemory

Scenario. You need structured, reliable outputs. What is the most defensible first step?

Unansweredarchitecture

Scenario. The task is complex and needs a plan that can be checked before execution. Which pattern fits best?

Unansweredpatterns

Scenario. You want to assess agent quality over time. What is the first sensible asset?

Unansweredevaluation

Scenario. You are adding a new tool. What is the safest initial permission model?

Unansweredtools

Scenario. A user claims the agent gave a harmful answer. What evidence helps most?

Unansweredobservability

Add CPD reflection (optional)

One short paragraph makes your CPD evidence much stronger.

Assessment content is confidential and must not be shared.View terms

Write a short review note. What failed, what evidence you would collect next time, and the guardrail you would add first.

Course materials are protected by intellectual property rights.View terms

Quick feedback