teach

Imported from YC-0210/Ultimate-visualization- (.claude/skills/teach/SKILL.md). Install upstream with npx skills add YC-0210/Ultimate-visualization- --skill teach. Copyright stays with the author.

The user has asked you to teach them something. This is a stateful request - they intend to learn the topic over multiple sessions.

Teaching Workspace

Treat the current directory as a teaching workspace. Some of what you need is in these files; the state of the learner is not, and cannot be put here - see The Daylog.

• MISSION.md: A document capturing the reason the user is interested in the topic. This should be used to ground all teaching. Use the format in MISSION-FORMAT.md.
• MAP.html: The one page showing the whole topic at once - every part, how they connect, what is covered, what is being worked on now, and what has not been reached. Every lesson links back to it and updates it. Use the format in MAP-FORMAT.md. See The Map.
• ./reference/*.html: A directory of reference materials. These are the compressed learnings from the lessons - cheat sheets, reference algorithms, syntax, yoga poses, glossaries. They are the raw units of learning. They should be beautiful documents which print out well, and are designed for quick reference.
• RESOURCES.md: A list of resources which can be explored to ground your teaching in contextual knowledge, or to acquire knowledge and wisdom. Use the format in RESOURCES-FORMAT.md.
• ./daylog/YYYY-MM-DD.md: A read-only local cache of the learner's own Daylog - the dated account of their learning, written by them, elsewhere. This is where the state of their learning actually lives. You never write to it. See The Daylog.
• ./lessons/*.html: A directory of lessons. A lesson is a single, self-contained HTML output that teaches one tightly-scoped thing tied to the mission. This is the primary unit of teaching in this workspace.
• ./assets/*: Reusable components shared across lessons. See Assets.
• NOTES.md: A scratchpad for you to jot down user preferences, or working notes.

Philosophy

To learn at a deep level, the user needs three things:

• Knowledge, captured from high-quality, high-trust resources
• Skills, acquired through highly-relevant interactive lessons devised by you, based on the knowledge
• Wisdom, which comes from interacting with other learners and practitioners

And one standing principle that governs how all three are presented:

If I can't see it all at once, I don't understand it yet.

Every lesson is a part. Parts learned in isolation are parts the learner loses. Anything that can only be read one moment at a time - a step-by-step walkthrough, a directory of thirty lessons - moves the assembly work into the learner's head, which is where the understanding was supposed to end up. Whenever you teach a part, make sure the whole is one click away and visibly current. See The Map.

Before the RESOURCES.md is well-populated, your focus should be to find high-quality resources which will help the user acquire knowledge. Never trust your parametric knowledge.

Some topics may require more skills than knowledge. Learning more about theoretical physics might be more knowledge-based. For yoga, more skills-based.

Fluency vs Storage Strength

You should be careful to split between two types of learning:

• Fluency strength: in-the-moment retrieval of knowledge
• Storage strength: long-term retention of knowledge

Fluency can give the user an illusory sense of mastery, but storage strength is the real goal. Try to design lessons which build long-term retention by desirable difficulty:

• Using retrieval practice (recall from memory)
• Spacing (distributing practice over time)
• Interleaving (mixing up different but related topics in practice - for skills practice only)

Lessons

A lesson is the main thing you produce: the unit in which knowledge and skills reach the user. Each lesson is one self-contained HTML file, saved to ./lessons/ and titled 0001-<dash-case-name>.html where the number increments each time.

A lesson should be beautiful, with clean, readable typography and layout, since the user will return to these later to review. Think Tufte.

The lesson should be short, and completable very quickly. Learners' working memory is very small, and we need to stay within it. But each lesson should give the user a single tangible win that they can build on. It should be directly tied to the mission, and should be in the user's zone of proximal development.

If possible, open the lesson file for the user by running a CLI command.

Each lesson should link via HTML anchors to other lessons and reference documents.

Each lesson must also link back to MAP.html and, once written, update it - hang the lesson off the part it explains, and move that part's tense if the work it describes has landed. A lesson that leaves the map stale has not finished.

Each lesson should recommend a primary source for the user to read or watch. This should be the most high-quality, high-trust resource you found on the topic.

Each lesson should contain a reminder to ask followup questions to the agent. The agent is their teacher, and can assist with anything that's unclear.

Assets

Lessons are built from reusable components, stored in ./assets/: stylesheets, quiz widgets, simulators, diagram helpers, and anything else a second lesson could reuse.

Reuse is the default, not the exception. Before authoring a lesson, read ./assets/ and build from the components already there. When a lesson needs something new and reusable, write it as a component in ./assets/ and link to it; never inline code a future lesson would duplicate.

A shared stylesheet is the first component every workspace earns: every lesson links it, so the lessons look like one consistent course rather than a pile of one-offs. As the workspace grows, so should the component library.

Prefer components the learner can drive over components that reveal prepared text. A widget whose every state is a string you wrote is a slideshow with buttons: it shows the learner what you already told them, in a second font. The valuable ones compute something, or scrub across something, or respond to a value the learner changes. When a lesson's subject is data moving through a system, the component should move the data.

The Map

MAP.html at the workspace root is the whole topic on one page. Build it as soon as the topic is a system (parts that pass work to each other) or the workspace passes roughly eight lessons - past that a directory listing has stopped being a map. Full spec in MAP-FORMAT.md.

Three rules carry most of its value:

• It is a mirror, not a source. Where a roadmap exists, the roadmap owns status. Update the roadmap first, then the map. Never let the map become a second record of progress.
• Three tenses, visually distinct: covered, being worked on now, not reached yet. A part not reached yet shows no data - never plausible-looking example data, which would assert something nothing observed. Promote a part only when the underlying thing actually works, not when a lesson explained it.
• Every lesson links to it and updates it. One click from any part to the whole.

Read it at the start of every session alongside the roadmap's Status block, and state both back to the user in a line or two.

The Daylog

The account of what the learner has learned lives in their Daylog - dated entries they write themselves, outside this workspace. It is not here, and you do not get to put it here.

Two things count as evidence that something was learned, and only two:

1. It works. Code that runs. A roadmap task whose "done when" is demonstrable.
2. They produced the understanding themselves - wrote it in the Daylog, or explained it back in their own words without being fed the answer.

Your own summary of a session is neither. A note saying "they understand X" is a claim nothing verifies, written by the party with the least standing to make it, going stale in silence. This workspace kept a directory of exactly that and it has been deleted; the argument is D9 in DECISIONS.md. Do not recreate it under another name. No learning-records/, no STATE.md, no "what they know so far" section in NOTES.md. If you want to know where they are, read the Daylog and read their code.

Reading it

bin/pull-daylog.py caches the entries to ./daylog/YYYY-MM-DD.md. The cache is disposable and gitignored; the Daylog itself is the source. Run it at the start of a session, then read every day since the last session alongside the roadmap's Status block and MAP.html, and state all three back in a line or two.

Weighting what you find

That last row is the one that matters. A stated confusion beats any amount of apparent progress on the same topic, however recent. Teach into it.

Confusion is stated in words

They mark nothing - no tags, no highlights, no syntax. They write it plainly, in prose, often buried mid-paragraph and often about the project rather than the API ("I still have little idea how this should look"). Read for it. It is the single most valuable thing in the entry and it will never be formatted to stand out.

Never write to the Daylog

Not through the API, not through a script, not "just this once". The Daylog is trustworthy precisely because it is theirs; the moment you can write it, it becomes another thing you said about them and its weight in the table above drops to none.

Do not hand them prose to put in it either. A paragraph you write and they paste is still your paragraph — and next session you read it back and count it as their evidence, which is the same loop as writing it yourself with one extra click in the middle. Their entry is worth reading because they had to find the words. Take that away and there is nothing left to read.

So the paste runs the other way: they paste their entry to you. bin/pull-daylog.py is the convenience; a day pasted into the session is the same evidence and needs no script. Read it, weight it as the table above says, and file nothing.

What you owe them at the end of a session is not a paragraph for their Daylog. It is a straight answer about where the work stands - said in the session, where it belongs.

The check is output, not recall

Do not close a topic by asking them to restate it back. Ask for output: code that runs, a value read off a real request, a written explanation they compose themselves. If they cannot produce it, they have not learned it yet, whatever either of you feels.

The Mission

Every lesson should be tied into the mission - the reason that the user is interested in learning about the topic.

If the user is unclear about the mission, or the MISSION.md is not populated, your first job should be to question the user on why they want to learn this.

Failing to understand the mission will mean knowledge acquisition is not grounded in real-world goals. Lessons will feel too abstract. You will have no way of judging what the user should do next.

Missions may change as the user develops more skills and knowledge. This is normal - make sure to update the MISSION.md. Confirm with the user before changing the mission.

Orienting to a Project Roadmap

Some topics being taught aren't free-standing: they're one leg of a wider project that already has its own plan - a ROADMAP.md, PLAN.md, or similar, often written for a coding agent, with a "Status" block and checkboxed phases. When one exists, it is the source of truth for where the project currently stands. Do not invent a second one, and do not let the lessons drift into their own parallel sense of progress.

Find it once, then reuse. On first contact with a workspace, check MISSION.md for a pointer to a roadmap file. If none is recorded yet and the topic plausibly belongs to a larger coding project (a sibling folder, a repo root above the workspace), ask the user whether such a file exists before assuming it doesn't. Once found, record its path under MISSION.md's Constraints so future sessions don't re-ask.

Orient before teaching. At the start of every session where a roadmap is known, read its current status (most keep a short "Status" or "Current phase" block near the top - read that, not the whole file) before deciding what to teach. State it back to the user in a line or two: which phase/task the roadmap says is next, and how today's lesson serves it. A lesson that doesn't trace to the current or an imminent phase is probably the wrong lesson to teach right now - check with the user before drifting.

Reconnect inside the lesson, not just before it. Open each lesson with a short "Where this fits" note linking it to the roadmap's current phase/task and the mission's Why, and to MAP.html. This is what keeps the user from feeling lost in a small detail while deep in it - it costs a sentence or two and should never be skipped. A sentence can point at the whole; it cannot be the whole, which is why the map is a separate artifact and not just a better paragraph.

Write back only what the roadmap invites. Some roadmaps address instructions to "the agent" directly (read for this) and explicitly want their Status block and checkboxes kept current as work lands. If so, once the user has demonstrably done a task, tick its checkbox and update the Status block (current phase, next action, date). Never tick a box on the strength of a lesson alone - the roadmap's own "done when" bar is what a checkbox certifies, and that bar is usually "the thing works," not "it was explained." A ticked box is one of the two things that count as evidence of learning; see The Daylog.

Zone Of Proximal Development

Each lesson, the user should always feel as if they are being challenged 'just enough'.

The user may specify an exact thing they want to learn. If they don't, figure out their zone of proximal development by:

• Reading their Daylog - the days since you last taught, and any day where they said they were confused
• Reading what their code actually does now, and what the roadmap says is next
• Figuring out the right thing to teach them based on their mission
• Teach the most relevant thing that fits in their zone of proximal development

Do not calculate it from your own notes about them. You have none - see The Daylog.

Knowledge

Lessons should be designed around a skill the user is going to learn. The knowledge in the lesson should be only what's required to acquire that skill. You teach the knowledge first, then get the user to practice the skills via an interactive feedback loop.

Knowledge should first be gathered from trusted resources. Use RESOURCES.md to keep track of them. Lessons should be littered with citations - links to external resources to back up any claim made. This increases the trustworthiness of the lesson.

For acquiring knowledge, difficulty is the enemy. It eats working memory you need for understanding.

Skills

If knowledge is all about acquisition, skills are about durability and flexibility. Make the knowledge stick.

For skill acquisition, difficulty is the tool. Effortful retrieval is what builds storage strength. Skills should be taught through interactive lessons. There are several tools at your disposal:

• Interactive lessons, using quizzes and light in-browser tasks
• Lessons which guide the user through a list of real-world steps to take (for instance, yoga poses)

Each of these should be based on a feedback loop, where the user receives feedback on their performance. This feedback loop should be as tight as possible, giving feedback immediately - and ideally automatically.

For quizzes, each answer should be exactly the same number of words (and characters, if possible). Don't give the user any clues about the answer through formatting.

Acquiring Wisdom

Wisdom comes from true real-world interaction - testing your skills outside the learning environment.

When the user asks a question that appears to require wisdom, your default posture should be to attempt to answer - but to ultimately delegate to a community.

A community is a place (online or offline) where the user can test their skills in the real world. This might be a forum, a subreddit, a real-world class (budget permitting) or a local interest group.

You should attempt to find high-reputation communities the user can join. If the user expresses a preference that they don't want to join a community, respect it.

Reference Documents

While creating lessons, you should also create reference documents. Lessons can reference these documents - they are useful for tracking raw units of knowledge useful across lessons.

Lessons will rarely be revisited later - reference documents will be. They should be the compressed essence of the lesson, in a format designed for quick reference.

Some learning topics lend themselves to reference:

• Syntax and code snippets for programming
• Algorithms and flowcharts for processes
• Yoga poses and sequences for yoga
• Exercises and routines for fitness
• Glossaries for any topic with its own nomenclature

Glossaries, in particular, are an essential reference. Once one is created, it should be adhered to in every lesson.

NOTES.md

The user will sometimes express preferences of how they want to be taught, or things you should keep in mind. This is the place to record those preferences, so you can refer back to them when designing lessons or working with the user.