How do I work with the AI tutor, and what are the limits?

Where the tutor is

The tutor panel sits inside the practicum lessons, next to the work you are doing. Coached build labs are the exercises where you write something (a prompt, a spec, a workflow) and get AI feedback on it. Foundations and tool modules have no tutor for anyone; they are cards you read.

Before your first message

The first time you use an AI feature you will see a short card, "Before you use AI feedback". It asks you to confirm three things:

• You will use hypothetical or sanitised facts only. Do not enter privileged, client-confidential, or ABA Rule 1.6 material.
• You understand that feedback is generated by an AI model and is not reviewed by a human before you see it.
• You have read what happens to what you type, in the AI Disclosure.

A short reminder stays next to every AI input afterwards.

Getting good answers

• Ask about the lesson you are in; the tutor is grounded in it.
• Paste the thing you built and ask what is missing, what would break, and why.
• Ask for the reasoning, not just the verdict.
• Keep a message under 2,000 characters. Long submissions are refused, so split them.

The limits

• Paid accounts: 200 AI interactions a day, counted across the tutor and coached feedback together. The counter resets at midnight UTC.
• Free accounts: the tutor and coached labs are available in the showcase module only, up to 30 a day each, resetting at midnight UTC.
• Each lesson conversation holds up to 200 messages.
• Checkpoints allow 10 attempts per quiz per day.

When you hit a limit the tutor says so; your lessons, exercises, and saved work stay open.

If you think the AI got it wrong

Email [email protected] with the lesson and what you submitted. A human reviews it.