Build a skill. Bring it to class.

1. Read an introductory provider lesson.
2. Ask an approved assistant to explain a familiar concept; mark facts, guesses, and omissions.
3. Write a five-sentence explanation of when teacher judgment is needed.

Success check: Your guide names two useful applications and two limitations, with a verified example.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Choose an objective, grade band, and expected student work.
2. Write a prompt specifying context, task, constraints, and output format.
3. Compare two drafts and revise the prompt using a clear quality criterion.

Success check: The revised output matches your objective and requires fewer factual or instructional corrections.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Generate a short answer on a topic you know well.
2. List three checkable claims and locate reliable original sources independently.
3. Annotate supported, unsupported, and misleading statements; write a corrected answer.

Success check: Every retained claim has a source that actually supports it; invented references are removed.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Select a short text you have permission to use and identify its essential ideas.
2. Create two versions with different vocabulary support and sentence complexity.
3. Compare each with the original and add a shared comprehension question.

Success check: Both versions retain key facts and require the same core reasoning.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Choose one learning objective and two likely misconceptions.
2. Draft five questions with answers and explanations using an approved planning tool.
3. Solve every question yourself; revise ambiguous distractors and plan a reteaching response.

Success check: Each item measures the objective and every distractor has an interpretable purpose.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Write a fictional student paragraph and a short rubric.
2. Request feedback tied to rubric evidence, with one actionable next step.
3. Remove unsupported judgments and compare the feedback against your own review.

Success check: Feedback points to specific text and helps revision without rewriting the work for the learner.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Choose a short explanation and identify barriers in its wording and layout.
2. Draft plain-language text and image descriptions; prepare an audio script with a transcript.
3. Review meaning and test reading order, keyboard navigation, and descriptions with a colleague.

Success check: Essential information is available without relying only on sight or sound.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Choose five subject terms and a short set of classroom directions.
2. Draft a bilingual glossary and examples using an approved tool.
3. Ask a proficient speaker to check meaning and register; retain the original text alongside the adaptation.

Success check: Translations preserve the concept, and learners can connect each term to an example.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Explore Microsoft’s Reading Coach training module.
2. Try a practice passage yourself and plan how to observe accuracy, expression, and understanding.
3. Design a brief follow-up discussion and an equivalent offline activity.

Success check: Your plan combines tool feedback with comprehension checks and human observation.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Choose a narrow topic and define what a successful explanation looks like.
2. Configure a teacher-controlled activity to ask one question at a time and provide hints.
3. Preview as a learner: test a wrong answer, an unclear answer, and a request for the solution; revise before any student pilot.

Success check: The prototype prompts reasoning and your plan specifies when a teacher intervenes.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Explore Code.org’s AI curriculum and select an activity for your grade band.
2. Rehearse the activity, noting how changing examples affects predictions.
3. Add a discussion of whose examples are missing and an exit question on model limitations.

Success check: Learners can explain why examples affect predictions and identify one limitation.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?

1. Define one learning problem, a baseline task, and a measurable success criterion.
2. Agree on school approvals, accessibility, a non-AI option, and what non-sensitive evidence to collect.
3. Plan a small trial and a colleague review; record benefits, errors, teacher time, and a continue/change/stop decision.

Success check: The decision uses learning evidence and review effort, not just engagement or novelty.

Reflect: What did you change after reviewing the AI output, and what would you do differently next time?