Best Practices
The AI assistant works best when you give it a clear goal and steady feedback. These practices help you get labs that are accurate and close to what you want, with fewer turns.
Write a clear prompt
Section titled “Write a clear prompt”The prompt is your main instruction. Aim for a specific, outcome-focused description:
- State the learning goal. Describe what a user should be able to do by the end, not just the topic. “Teach a developer to deploy a service and roll back a bad release” works better than “a lab about deployments”.
- Name the scenario. Mention the tools, product, or workflow the lab centers on.
- Add hard requirements. Say what the assistant must include or avoid — specific commands, a particular flow, concepts to skip.
- Keep it focused. One objective per lab gives a tighter result than a prompt that tries to cover everything.
- Show, don’t only tell. Attach a screenshot or diagram when it explains what you want faster than words.
Choose settings that match your audience
Section titled “Choose settings that match your audience”- Duration sets the scope. A 10-minute lab is a focused exercise. A 2-hour lab is a deeper, multi-chapter journey. Longer labs get more chapters, not just more text, so match the duration to how much you want to cover.
- Audience experience sets depth and pacing. Pick Beginner for step-by-step fundamentals. Pick Advanced when users already know the basics and you want to move faster.
Review as you go
Section titled “Review as you go”- Check the first chapter carefully. It sets patterns that the rest of the lab follows. Fix it early, before the same issue repeats.
- Give specific feedback. When something is wrong, say exactly what to change. “Use our CLI instead of the web console” gets a better correction than “this isn’t right”.
- Answer questions directly. Clear answers keep the assistant on track.
- Test often. After a turn, click Test lab to have the assistant play through the lab and report what breaks, or Play to try it yourself.
Iterate on a lab
Section titled “Iterate on a lab”A lab the assistant builds is a strong starting point, not a finished product:
- Ask for one change at a time. A specific request (“rename chapter 1 to Getting Started”) lands more reliably than a list of loosely related ones. It also costs less.
- Make small edits yourself in the lab editor when that is faster than describing them. Your edits and the assistant’s share the same working branch.
- Restart a session that has drifted. When the assistant keeps working from old context, restart the session and restate what matters.
- Always review before you publish. Read and test the lab with the same care you give hand-written content before it reaches users.