AI agents practice assessment. Core concepts

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. You split a task into planner and executor agents. What must you add first?

Unansweredmulti-agent

Scenario. A tool call fails intermittently. What is the most defensible response?

Unansweredtools

Scenario. The agent retrieves the wrong document. What should you improve first?

Unansweredretrieval

Scenario. A tool grants refunds. What is the safest first guardrail?

Unansweredsecurity

Scenario. You need a machine readable plan. What should you enforce?

Unansweredoutputs

Scenario. Costs spike during peak traffic. What is the cleanest control?

Unansweredcost

Scenario. Users want to know why the agent said something. What is the best signal to show?

Unansweredquality

Scenario. The agent fails to meet its objective. What should the system do next?

Unansweredresilience

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