Tutor notes as a course tree: AI drills, local truth

You teach the same topic three times a week and still lose the “good” version of an explanation in a chat scroll or a random Word file named final_v7. Students get uneven handouts. You invent quiz variants from scratch at midnight. The fix is not another AI tutoring product — it is a course tree you own, with models used only for practice generation and wording help.

Three builds ship from this site. SEO Note 2.1 is the current one: an Electron app (about 121 MB installer) for Windows 10/11 64-bit and Linux x64, with Markdown, the tree editor, AI talking to a model you connect — a local one or any OpenAI-compatible endpoint, a web clipper and git sync of the file. For Android 6.0 and newer there is a release-signed APK (~10 MB) of the same interface; on the phone notes live as a folder of Markdown files on the device, and passwords, sync, the clipper and AI answer there with an explicit refusal. The classic build 1.3.34 (~2.5 MB, Win32) stays on the download page for Windows 7, 8 and 32-bit PCs, where 2.1 does not install; its notes file is .seo. All three work offline with no account: pick your build, or read how 2.1 and 1.3.34 differ.

SEO Note is a free Windows tree outliner (one local file, no account). It does not grade, sync, or call models. Your hierarchy holds Course → topic → assignment → question bank; AI stays in the browser as a temporary helper on a slice you paste. That split matches the wider habit in local notes as source of truth for AI.

Where tutoring work lives

In SEO Note

• Syllabus spine
• Verified explanations
• Assignment specs
• Question bank + rubrics

In the AI chat

• Extra drill variants
• Plain-language rewrites
• Analogy hunting
• Disposable brainstorms

If a generated quiz never returns to the Question bank node, it was a one-off — fine for tonight, useless next semester.

Course tree that survives a semester

Tutoring
  Math — grade 9
    Topics
      Fractions — mixed numbers
        Core explanation (your words)
        Common mistakes
      Linear equations
    Assignments
      HW-04 — due 2026-03-12
        Spec + constraints
        Rubric bullets
    Question bank
      Easy / Medium / Hard
      Oral warm-ups
    AI scratch (dated)
  English — IELTS
    Topics
    Assignments
    Question bank

One parent per course keeps search calm. Put the explanation you actually say aloud under Core explanation — not a paste of a textbook chapter. Common mistakes is the highest-yield branch before a retest.

What to paste into AI

Paste a thin slice: one topic’s Core explanation + three Common mistakes + the skill you want practiced. Ask for drills, not for “teach this course.” Keep student names and scores out of the window.

Prompt (reuse under AI scratch):
You are a drill generator, not a grader.
Use ONLY the notes below. Do not invent course policy.
Produce 5 short practice items (Easy→Hard) + one oral warm-up.
Mark any assumption you had to make.
---
[paste Core explanation + Common mistakes]

Do not paste: student full names with contacts, gradebooks, parent emails, passport/ID scans, or whole LMS exports. AI helps with drills — it is not your gradebook or CRM.

How to set this up in one sitting

1. Create Tutoring → one course → Topics / Assignments / Question bank / AI scratch.
2. Pick the topic you teach this week. Write Core explanation in your voice (ten lines is enough).
3. Add three Common mistakes you already heard in class.
4. Paste that slice into AI with the prompt above; keep the chat open only until you pick winners.
5. Rewrite accepted items into Question bank under Easy / Medium / Hard — your wording, not a raw dump.
6. Link the week’s Assignment node to the bank items you will actually use; date the due field.
7. Close the chat. Next week starts from the tree, not from “that conversation somewhere.”

From class to bank

1 Capture Core explanation

2 Log mistakes heard live

3 Generate drills on that slice

4 Promote winners into the bank

5 Grade from rubric in the file

Grading stays local

Rubric bullets live under the Assignment node. When AI suggests “score 8/10,” treat it as noise unless you already decided the criteria. Models flatten nuance and invent fairness. Your rubric is the authority — same spine as source-of-truth notes, applied to teaching.

• Oral warm-ups: keep three lines max; regenerate variants only from those lines
• Retakes: clone Hard items and change numbers by hand so memory of last week’s sheet does not carry over
• Shared materials for a co-teacher: export or copy a branch; do not rely on a chat thread as the handoff

	Chat as curriculum	Tree + AI drills
Next semester	Hunt old threads	Reuse Question bank
Fairness	Model “scores”	Your rubric node
Student privacy	Easy to overshare	Paste topics, not dossiers
Offline / lab Wi‑Fi	Tutor UI gone	File still opens

Pitfalls and honest limits

• Letting AI write the Core explanation first — then you never sound like yourself in class
• Dumping an entire textbook PDF into chat “for context” — noisy, often against policy, and you stop curating
• Storing grades only in a model memory feature — vendor pivot, access loss, privacy mess
• SEO Note will not schedule lessons, host video, or sync to a phone (the 2.1 APK keeps its own Markdown folder, no sync); pair with whatever calendar you already use
• No spaced repetition, no LMS hooks and no model in the box — 2.1's AI calls an endpoint you connect, the APK has none; Question bank nodes are intentional lists, not algorithms
• Live co-authoring of one syllabus with a whole department belongs in a shared doc; keep your private bank and mistake log local

Students running their own exam tree can mirror this pattern from the other side — see study notes with AI. Freelancers who tutor as a side business often keep client delivery separate from course content; freelancer notes tree covers that split. For a method lens on Areas vs Projects, PARA in a notes tree sits nearby. Ready-made skeletons: templates.

Next steps

1. Create one course parent with Topics, Assignments, Question bank, AI scratch
2. Fill Core explanation + three mistakes for this week’s topic
3. Run one drill-generation pass; promote five items into the bank
4. Attach a short rubric under the active Assignment
5. Copy the notes file to USB or another disk before the busy exam week

Download SEO Note free and keep the course in a file you control — models can invent practice; only you own the grade truth.

Keep going

• Tree templates and PDF guide — role and method skeletons you can copy straight into your own tree.
• If this guide helped, you can say hello or grab ready-made trees on the templates page.

Adjacent tools (VPS for local models, courses, VPN) are not stuffed into every paragraph. If an affiliate link appears, it will be labeled and is never required to use SEO Note.

Mini FAQ

How big is the installer for the tutor course tree? About 121 MB for 2.1, ~2.5 MB for the classic 1.3.34 — one local file, no account - the app does not grade, sync, or call models.

Can it grade student work or decide a score? Grading stays with your rubric under the Assignment node; a model suggesting a score is noise unless you already decided the criteria.

What student details must never go into the chat? Full names with contacts, gradebooks, parent emails, passport or ID scans and whole LMS exports - AI helps with drills, not as a gradebook or CRM.

How long should a core explanation note be? Ten lines written in your own voice is enough for the topic you teach this week, with three common mistakes added beside it.

Does it sync lesson plans to a phone or run video? It will not schedule lessons, host video, or sync to a phone — the 2.1 APK keeps its own Markdown folder on the device — so pair it with whatever calendar you already use.

Related: Source of truth for AI · Study notes + AI · Freelancer notes tree · PARA method tree · Templates · FAQ