Are you ready to build agentic AI?Free Agentic AI skill check.
The Agentic AI skill check is five multiple-choice questions, one each on Tool use, Agent loops, Planning, Memory and Agent safety. It is free, takes about three minutes, and grades you from Novice to Expert with every answer explained.
- 5 questions
- About 3 minutes
- Free
Choose a course
Every course in the AI Tutor has its own check: five questions, one per topic.
Career tracks 11 courses
For a new role: engineer, scientist, analyst or builder.
Sign up free to start. Your answers are graded when you finish and saved to your record.
The Agentic AI skill check.Five topics, one question each.
Question 1 is below exactly as the check asks it. The answer, and why it is right, come with your result.
Agentic AI is the shift from models that answer questions to systems that take action — calling tools, making decisions, and completing multi-step tasks on their own. It's one of the fastest-moving areas in AI, and the engineers who can design reliable agents are in short supply.
This quick diagnostic tests five foundations that separate a working agent from a demo: tool use and function calling, the reason–act loop that drives an agent, planning and task decomposition, memory systems for state and recall, and the safety concerns unique to systems that can act.
It takes about 3 minutes with a free student account, and gives you an instant read on where you stand before diving into the full Agentic AI track.
- Tool use
Calling tools is what turns a language model into an agent that can actually act.
- Agent loops
The reason–act–observe loop is the core control flow of every autonomous agent.
- Planning
Decomposing a goal into ordered steps is what lets agents tackle complex, multi-step tasks.
- Memory
Memory systems give agents state and recall beyond a single context window.
- Agent safety
Agents with tool access are high-risk for prompt injection: safety is non-negotiable.
In an agentic system, 'tool use' (function calling) lets the model…
- ARun with no prompt at all
- BRetrain itself between messages
- CCall external functions or APIs and use their results to act
- DEnlarge its own context window
“Learn AI” means three things.This check tests one of them.
It tests building agents on LLMs: tool use, agent loops, planning, memory and guardrails.
- 01
Machine learning foundations
How models learn from data: the statistics, training and evaluation underneath them.
For you ifyou want to train and judge models, as an ML engineer or data scientist.
Take the Machine Learning check - 02This check
Applied LLM engineering
Building software on models like Claude and GPT: prompts, retrieval, agents, evals and shipping.
For you ifyou write code and want to build AI features, not train models from scratch.
- 03
AI for your job
Using AI tools well in the work you already do: briefing them, checking what comes back, keeping data safe.
For you ifyou do not plan to code and want AI to make the work you do faster and better.
Take the AI for Operations check
Three steps.About three minutes.
It is free and nothing is booked. Save the report to keep it on your account.
- 01
Pick a course
Any of the 21 courses in the AI Tutor, from Generative AI to AI for Teachers. Each check covers five topics.
- 02
Answer five questions
One question per topic, four options each. Nothing is marked until you finish.
- 03
See your readiness
Your readiness out of 10, a skill map, every answer explained, and the path through the course.
Programmes that build on this check.Your result shows where to start.
Each one teaches Agentic AI by building, and ends in deployed work graded against a rubric you can read.
Before you start.Three quick answers.
What does an Agentic AI engineer do?
They design systems where a language model autonomously uses tools, plans multi-step work, keeps memory, and completes tasks — from coding agents to research and workflow automation.
How is this different from the LLM Agent Architect check?
The foundations overlap because both build on agent fundamentals. The full Agentic AI course focuses on building and shipping agents end-to-end; LLM Agent Architect goes deeper into protocols, harnesses, and frameworks.
Do I need to code to take it?
No — this quick check tests conceptual understanding. The full course teaches you to build agents hands-on, so some programming experience helps there.
Rather talk to a person?
Tell us what you want to do with AI. We’ll suggest a programme on a 15-minute call.