AI Agents · Course resources

Notes: Run Your First AI Agent Without Writing Code

Review the key ideas from Run Your First AI Agent Without Writing Code.

From watching to trying

In the opening comparison, Claude Code chose searches and gathered prices and guest-access conditions from different sources. You supplied the goal; the AI chose some steps and used tools to carry out the research.

On Your Page, keep Your first explanation and Compare the explanations. Your first sentence gives you something to compare with your explanation at the end of the course.

Check what a product actually does

A product that drafts a customer reply may still leave someone to look up the information and send the message. Compare the advertised job with the actions shown. Exaggerating a product’s agent capabilities is called agent-washing.

The four course stages describe what you can do: Nodding means recognizing the term; Naming means explaining the work; Running means trying a job and checking its result; Directing means choosing the work, action boundaries and checks.

Try a research comparison

First check one current price. Inspect the activity, open the source and compare the plan name, billing period and price. Record the question, source and any uncertainty under Your price check on Your Page.

Use The wording for your first comparison to give an agent a decision and three requirements. The resource includes setup links. Use the normal default settings, ask for a table with sources and check the claims that affect your decision.

Under Your first comparison, keep your decision, request, result and any missing information or unfinished work. These notes let you compare this first attempt with later practice.