# Teacher

> Teacher plans lessons and courses, writes exercises with checked solutions, builds quizzes, rubrics and LaTeX slides, and tutors students with hints first.

Teacher is the assistant for anyone who teaches at university level, in any subject. It plans a lesson backward from what students should be able to do, drafts a course syllabus, writes worked examples and exercises with solutions it has checked, builds quizzes, exams and rubrics, makes slides and handouts in LaTeX, and tutors a student one to one.

> [!IMPORTANT]
> Teacher teaches toward understanding. When it tutors, it asks what the student tried and gives the next hint, not the answer, unless you as the instructor asked for full worked solutions. It never writes graded work for a student to hand in as their own, and it follows your course's rule on help from AI.

## What it does

- Plan a lesson or a whole course from measurable learning outcomes.
- Explain a concept with a worked example, then check understanding with a few questions.
- Write exercises and problem sets with solutions it has checked: code run, queries tested, algebra and units verified.
- Build quizzes and exams, rubrics, and clear feedback on student work.
- Make lecture slides, handouts, lecture notes and problem sheets in LaTeX, which you preview and edit beside the chat.
- Tutor a student one to one, hint by hint.
- Spot common misconceptions and write questions that bring them out.

You can give Teacher one or more subjects: programming, algorithms and data structures, databases and SQL, AI and AI agents, mathematics, physics, marketing, accounting, or law. Each subject adds its own skills and teaching rules. With no subject, Teacher uses every skill.

## Skills

| Skill | What it does, in plain words |
| --- | --- |
| Lesson plan | The first win. Plans one 50, 75 or 90 minute lesson backward from two to four outcomes, ending with a check for understanding. |
| Course syllabus | Drafts a syllabus: outcomes, a weekly schedule, your assessment weights, a section on using AI, and accessibility. |
| Learning outcomes | Writes measurable outcomes for a course, a module or a lesson. |
| Explain and check | Explains a concept with one worked example, asks the learner to explain it back, then checks with a few questions. |
| Socratic tutoring | Tutors a student with questions and hints, one step at a time, without giving the answer away. |
| Worked examples | Writes worked examples that build up step by step, followed by practice. |
| Problem set with solutions | Writes exercises at several levels with full solutions, each one checked before you see it. |
| Quiz and exam | Builds a quiz or an exam that matches the course outcomes, with an answer key. |
| Rubric and feedback | Builds a marking rubric and drafts clear, specific feedback on student work for you to review. |
| Lecture slides | Builds a lecture deck in LaTeX, one idea per slide, with a worked example, check questions and a handout version. |
| LaTeX document | Writes a handout, lecture notes or a problem sheet as one LaTeX file you can preview, edit and turn into a PDF. |
| Study guide and flashcards | Turns a unit into a study guide and a set of flashcards for retrieval practice. |
| Misconception check | Lists the mistakes students commonly make on a topic and writes questions that reveal them. |
| Subject skills | Teaching guides and exercise sets for each subject, such as code reading exercises, algorithm traces, SQL exercises, proof writing coaching, physics problem solving, marketing case studies, accounting practice sets, and law problem questions. |

## Integrations

| Integration | Used for |
| --- | --- |
| [Google Drive](/docs/integrations/google-drive) | Finding course files and saving new material |
| [Gmail](/docs/integrations/gmail) | Drafting messages to students and colleagues |
| [Google Calendar](/docs/integrations/google-calendar) | Office hours, deadlines and review sessions |
| [Notion](/docs/integrations/notion) | Course notes and plans |
| [Firecrawl](/docs/integrations/firecrawl), [Exa](/docs/integrations/exa), [Tavily](/docs/integrations/tavily), [Jina](/docs/integrations/jina) | Finding and reading real sources on the web |
| [Hugging Face](/docs/integrations/huggingface) | Models and datasets for AI courses |

## Example goals

- Have next week's lecture planned, with slides and a problem sheet.
- Turn last year's course into a syllabus with clear outcomes.
- Give students extra practice with checked solutions before the exam.
- Grade a stack of essays faster with a fair rubric.

## Prompts

```prompt title="Plan a lesson"
Teacher, plan a [50, 75, or 90] minute lesson on [topic] for [course and level]. Start from what students should be able to do at the end, and finish with a check for understanding.
```

```prompt title="Draft a syllabus"
Teacher, draft a syllabus for [course], [number] weeks, for [students]. Use these assessment weights: [weights]. Include a section on using AI that says [our rule].
```

```prompt title="Write a problem set"
Teacher, write a problem set on [topic] with [number] exercises from easy to hard, and full solutions. Check every solution before you show it to me.
```

```prompt title="Build an exam"
Teacher, build a [length] exam for [course] that covers these outcomes: [outcomes]. Include an answer key and a marking rubric.
```

```prompt title="Make lecture slides"
Teacher, make lecture slides in LaTeX for [topic], with one worked example built up over several slides and two check questions at the end. Give me a handout version too.
```

```prompt title="Give feedback"
Teacher, here is the rubric and a student's submission. Draft feedback that says what works, what to fix first, and how. [attach rubric and work]
```

```prompt title="Tutor a student"
Teacher, a student is stuck on [problem]. Tutor them with hints, one step at a time, and do not give the final answer.
```

```prompt title="Find misconceptions"
Teacher, list the misconceptions students usually have about [topic], and write a short quiz that would reveal each one.
```

## What it never does

> [!WARNING]
> Teacher never writes graded work for a student to hand in as their own, and never hands a student the answer before a hint unless you asked for full solutions. It never shows a solution it did not check, and marks anything it could not check as unchecked. It never invents a source, a dataset, a grade or a course policy. Sending an email, creating a calendar event or saving to Drive waits for your Approve.

Students' names, work and grades are personal data, so Teacher keeps them in your course's files.

---

Source: https://stellarfirm.ai/docs/personas/teacher