Agent skill

Interactive Course Builder

by XiaomiMiMo in XiaomiMiMo/MiMo-Code

Turns a PDF, paper, document, URL or topic into a chapter-by-chapter course with exercises, feedback and review, and saves progress so you can resume later.

MITAuto-check passedEducation

Install Interactive Course Builder

skills CLI
$ npx skills add XiaomiMiMo/MiMo-Code --skill learn-everything -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install XiaomiMiMo/MiMo-Code learn-everything --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/XiaomiMiMo/MiMo-Code.git skills-src && mkdir -p .claude/skills && cp -r skills-src/packages/cli/src/skill/builtin/.bundle/learn-everything .claude/skills/learn-everything && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
learn-everything
GitHub stars
14k
Token cost
~2.9k tokens
SKILL.md length
1,547 words
Files
4 (incl. references)
Skills in repo
22
Repo updated
First seen
Licence
MIT

At a glance

Turns a PDF, paper, document, URL or topic into a chapter-by-chapter course with exercises, feedback and review, and saves progress so you can resume later.

  • Works in 7 steps: Read the document's structure first:… → Identify the table of contents,… → Use semantic units rather than equal… → …
  • Studying a PDF, paper or book chapter step by step
  • SKILL.md covers Core behavior, Determine the input mode, Start the course and Build the course map, plus 7 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

The agent builds an adaptive course that alternates explanation, retrieval, application, feedback and review. It shows a course map and then teaches one chapter or lesson at a time, in your language, without revealing exercise answers before you try them. It separates claims from the source from background knowledge it adds, and matches depth and notation to your level.

For documents it reads the structure first, reads short sources of about 30 pages or fewer in full before finalizing the map, and reads longer ones lazily, one chapter before teaching it. Units follow meaning, not equal page ranges, and source-specific teaching cites page numbers and sections, flags unreadable or contradictory content and checks figures and tables. Progress lives in a `course-state.md` file under a `learn/` folder, or in a copyable state block when file tools are missing. Reference notes cover practice and session formats.

When your agent uses it

  • Studying a PDF, paper or book chapter step by step
  • Getting a tutorial or curriculum from source material, with quizzes and hints
  • Continuing a previous course from a saved state file or pasted state block
  • Learning a topic from scratch with a structured course map

Example prompts

  • “Teach me this PDF chapter by chapter and quiz me after each section.”
  • “Build a course on the basics of Bayesian statistics and start with the course map.”
  • “Continue my course from the state block I pasted below.”
  • “Turn this paper into lessons, citing page numbers, and hold back the exercise answers.”

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Read the document's structure first: table of contents, headings, section lengths, and spot checks of dense sections. Fully read short…
  2. Identify the table of contents, headings, prerequisites, central claims, definitions, examples, equations, figures, tables, and appendices.
  3. Use semantic units rather than equal page ranges. Merge short dependent sections and split dense sections.
  4. Cite page numbers and section names when teaching source-specific content.
  5. Examine relevant figures, tables, and diagrams rather than relying only on extracted text.
  6. Mark unreadable, missing, contradictory, or ambiguous source content explicitly. Do not invent a reconstruction.
  7. Treat the supplied source as the primary curriculum. Add outside material only to supply prerequisites, clarify, correct an apparent…

What it can do on your machine

Read from SKILL.md and the folder at commit 6babeb0. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Interactive Course Builder loads about 2.9k tokens when it runs, and up to ~4.7k if it reads all its reference files. Until then it costs about 192 tokens; SKILL.md has 1,547 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~192
When it runs · the whole SKILL.md, loaded when a task matches
~2.9k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~4.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from XiaomiMiMo/MiMo-Code at commit 6babeb0, republished under its MIT licence (© XiaomiMiMo). 1,547 words, ~2,905 tokens.

Download SKILL.mdSave it as .claude/skills/learn-everything/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
learn-everything
description
Turn an uploaded PDF, paper, book chapter, document, URL, or user-provided topic into a structured, interactive learning course. Use when the user wants to learn, study, understand, master, review, or practice a subject chapter by chapter; asks for a tutorial or curriculum from source material; wants exercises, quizzes, answer grading, hints, spaced review, or a final assessment; or says "teach me this", "learn this PDF", "分章节教学", "带我学", or similar; or returns to continue a previous course ("continue my course", "接着上次学") or supplies a saved course-state file or state block. Adapt explanations and practice to the learner's level, preserve page or section references for documents, and teach incrementally rather than dumping all content at once.

Learn Everything

Convert source material or a topic into an adaptive course that alternates explanation, retrieval, application, feedback, and review.

Core behavior

  • Teach in the user's language unless they request another language.
  • Optimize for durable understanding, not merely summarization.
  • Work incrementally. Present a course map, then teach one chapter or lesson at a time.
  • Do not reveal exercise solutions before the learner attempts them, unless the learner explicitly asks.
  • Persist learning state. When file tools are available, maintain learn/<slug>/course-state.md (course map with completion, concept mastery table, review queue, error log) and update it at chapter boundaries — not every turn. Without file tools, keep state in conversation and emit a compact copyable state block at each chapter checkpoint so the learner can save it and paste it back to resume later. See references/session-formats.md for both formats.
  • Prefer concrete examples, analogies, diagrams described in text, counterexamples, and worked derivations over abstract exposition alone.
  • Separate source claims from added background knowledge. Never attribute added material to the source.
  • Match depth and notation to the learner. Define unfamiliar terms before relying on them.

Determine the input mode

Document mode

Use document mode for PDFs, papers, books, slides, notes, or other supplied material.

  1. Read the document's structure first: table of contents, headings, section lengths, and spot checks of dense sections. Fully read short sources (a paper, a chapter, roughly ≤30 pages) before finalizing the course map; for longer sources, finalize the map from structure and read full text lazily — only the chapter currently being taught, immediately before teaching it.
  2. Identify the table of contents, headings, prerequisites, central claims, definitions, examples, equations, figures, tables, and appendices.
  3. Use semantic units rather than equal page ranges. Merge short dependent sections and split dense sections.
  4. Cite page numbers and section names when teaching source-specific content.
  5. Examine relevant figures, tables, and diagrams rather than relying only on extracted text.
  6. Mark unreadable, missing, contradictory, or ambiguous source content explicitly. Do not invent a reconstruction.
  7. Treat the supplied source as the primary curriculum. Add outside material only to supply prerequisites, clarify, correct an apparent error, or provide practice; label it as supplemental.
Topic mode

Use topic mode when the learner supplies a subject rather than a source.

  1. Define a sensible scope and end capability.
  2. Identify prerequisites and arrange concepts from foundational to advanced.
  3. For current, niche, contested, or externally verifiable topics, research authoritative sources with available web tools before constructing the curriculum.
  4. Prefer primary sources and official documentation for technical or scientific claims.
  5. Distinguish stable fundamentals from current developments.
  6. State material scope and assumptions in the course map.

Do not delay progress for minor ambiguity. Choose a useful default scope and make it visible.

Resume mode

Use resume mode when the learner supplies a saved course-state file or state block, or asks to continue a previous course.

  1. Read the state: completed chapters, concept table, review queue, error log.
  2. Run 2–4 warm-up retrieval questions before new content, weighted toward review-needed concepts and past error classes. Ask fresh questions, never the originals verbatim.
  3. Give a one-line position recap, update the state with warm-up results, then proceed with the current chapter.
  4. If the source document is needed but not re-supplied, ask for it; teach from the state's concept summaries only as a stopgap and say so.

Start the course

Gather only information that materially changes the lesson. If the learner's level or goal is unknown, either:

  • ask one compact diagnostic question, or
  • offer a 3-question diagnostic quiz and let the learner skip it.

Do not ask a long questionnaire. Infer preferences from the conversation whenever possible.

Then provide:

  1. Target capability — what the learner should be able to explain or do.
  2. Course map — 3–12 chapters, each with purpose and prerequisites.
  3. Estimated level — beginner, intermediate, advanced, or mixed, with a brief reason.
  4. Learning contract — explain that lessons proceed one at a time and answers remain hidden until attempted.
  5. Next action — begin the first lesson or diagnostic immediately.

Use the detailed formats in references/session-formats.md when consistency is useful.

Build the course map

Create chapters around conceptual dependencies, not source length alone. Each chapter should have:

  • a concise title;
  • one observable learning objective;
  • key concepts;
  • prerequisite links;
  • source pages or sections in document mode;
  • one planned practice outcome.

Use these sizing rules:

  • Split a chapter when it contains more than one major reasoning jump.
  • Merge sections that cannot be understood independently.
  • Create a prerequisite bridge chapter when the source assumes missing knowledge.
  • Keep advanced appendices optional unless needed for the learner's goal.
  • For sources with more than ~12 natural chapters, group chapters into parts, agree on scope with the learner, and run the course part by part rather than flattening everything into one map.
  • Reorder only when pedagogically necessary; explain the reordering in document mode.

Teach each lesson

Follow this lesson loop:

  1. Orientation — state why the lesson matters and connect it to prior learning.
  2. Objectives — give 2–4 observable goals.
  3. Core explanation — explain in small sections, introducing one conceptual jump at a time.
  4. Worked example — demonstrate the reasoning, not only the final result.
  5. Misconception check — identify 1–3 likely confusions or boundary cases.
  6. Active recall — ask a brief check before proceeding to longer practice.
  7. Practice — assign exercises calibrated to current mastery.
  8. Feedback loop — grade the learner's response and decide whether to advance, remediate, or increase difficulty.

For dense material, pause after each major concept and ask a single check question. For straightforward material, teach the full lesson before practice.

Show full SKILL.md (619 more words)Show less

Design practice

Select exercises from multiple cognitive levels:

  • Recall: define, list, label, reproduce.
  • Explain: restate in the learner's own words; compare concepts; explain why.
  • Apply: solve a representative problem or use the method in context.
  • Debug: find an error in reasoning, code, derivation, or interpretation.
  • Transfer: apply the idea in a novel setting.
  • Synthesize: connect several chapters or produce an artifact.

Default to 3–5 exercises per lesson, mixing at least two levels. Prefer one exercise at a time when grading interactively. Use larger sets only when the learner requests a worksheet or exam.

Consult references/practice-and-mastery.md for exercise-writing rules, difficulty bands, adaptation, and grading rubrics. The cognitive levels above describe what the learner does; the difficulty bands in the reference describe how much scaffolding — they combine freely.

Grade and respond to learner answers

Evaluate substance before wording. Use this sequence:

  1. State the verdict: correct, mostly correct, partially correct, or needs revision.
  2. Identify what was done well.
  3. Point to the first important gap or error.
  4. Explain the correction with the smallest useful amount of teaching.
  5. Ask for a revision, provide a targeted follow-up, or advance.

Do not merely say “correct.” Reinforce the decisive reasoning.

When the answer is wrong:

  • First provide a hint that targets the misconception.
  • Give a stronger hint after another failed attempt.
  • Reveal a worked solution after repeated attempts, an explicit request, or when further guessing has low learning value.
  • Immediately follow a revealed solution with a near-transfer question.

When the answer is correct:

  • Increase difficulty gradually.
  • Ask for explanation if the learner may have guessed.
  • Use a transfer question when mastery appears strong.

Track mastery and review

Maintain a concept table — one row per concept — in the course state (file or state block):

  • status: unseen, introduced, practicing, mastered, or review-needed;
  • evidence: strongest recent answer or error;
  • confidence: low, medium, or high;
  • next review point.

Update it at chapter boundaries and after significant errors. Do not print the full table every turn; show a compact progress snapshot at chapter boundaries or when requested.

Schedule cumulative review:

  • after each chapter: 2–4 retrieval questions from current and earlier material;
  • after roughly one-third and two-thirds of the course: mixed checkpoint;
  • at course end: final assessment and synthesis task;
  • revisit recurring errors more often than mastered concepts.

Treat mastery as demonstrated transfer, not one correct recall answer.

Handle learner commands

Recognize and act on natural commands such as:

  • “继续 / continue” — move to the next pending step.
  • “提示 / hint” — give the next hint without the full answer.
  • “答案 / show solution” — reveal and explain the solution, then ask a near-transfer question.
  • “更简单 / simpler” — reduce abstraction, add prerequisites and examples.
  • “更深入 / deeper” — add derivation, edge cases, and harder transfer.
  • “跳过 / skip” — mark the item skipped and continue.
  • “复习 / review” — generate retrieval practice weighted toward weak concepts.
  • “考试 / test me” — run a closed-book mixed assessment.
  • “进度 / progress” — show completed chapters, mastery, weak points, and next step.
  • “重做课程 / rebuild course” — revise scope or chaptering while preserving demonstrated mastery.

Complete the course

At the end, provide:

  1. a mixed final assessment without solutions shown initially;
  2. a synthesis task that uses multiple chapters;
  3. grading with concept-level feedback;
  4. a concise mastery report;
  5. a personalized review plan focused on weak concepts;
  6. a compact reference sheet created only after assessment, so it does not replace retrieval practice.

For document mode, include a source navigation index linking mastered concepts back to pages or sections.

Quality checks

Before each response, verify:

  • Solutions are not exposed before the learner attempts the exercise.
  • The current activity matches the course state, and the state is updated at chapter boundaries.
  • The lesson does not assume an unexplained prerequisite.
  • The response gives the learner exactly one clear next action.

© XiaomiMiMo, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (references) in packages/cli/src/skill/builtin/.bundle/learn-everything of XiaomiMiMo/MiMo-Code.

  • SKILL.md
  • agents/openai.yaml
  • references/practice-and-mastery.md
  • references/session-formats.md

Open the folder on GitHubat commit 6babeb0

Compare with similar skills

Interactive Course Builder next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Interactive Course Builder compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Interactive Course Builder this skillXiaomiMiMo/MiMo-Code14k—~2.9kAutomated safety check: PassMIT
Claude Certification Tutorrohitg00/ai-engineering-from-scratch65k—~3kAutomated safety check: PassMIT
StudyVault Quiz Tutorbevibing/tutor-skills1.3k—~1.4kAutomated safety check: PassMIT
Codex Skill Self-AssessmentFlorianBruniaux/claude-code-ultimate-guide6.1k—~2.3kAutomated safety check: PassCC-BY-SA-4.0
Study Companion Enmingchen666/Reviva236—~5.7kAutomated safety check: PassNone
Learn Law With Rohasrohasnagpal/legal-ai-skills174—~2.5kAutomated safety check: PassMIT

Similar skills

  • Claude Certification Tutor

    rohitg00/ai-engineering-from-scratch

    Guides a learner through one of four independent Claude certification tracks with onboarding, lessons, practice labs, mock exams and remediation.

    65k GitHub stars~3k tokensUpdated today
    EducationAuto-check passed
  • StudyVault Quiz Tutor

    bevibing/tutor-skills

    Quizzes you on the notes in an Obsidian StudyVault, tracks proficiency per concept and drills weak areas in four-question rounds.

    1.3k GitHub stars~1.4k tokensUpdated 7 mo ago
    EducationAuto-check passed
  • Codex Skill Self-Assessment

    FlorianBruniaux/claude-code-ultimate-guide

    Runs an interactive quiz in a quick or comprehensive mode, scores your skill level by topic and generates a personalized learning path with practice projects.

    6.1k GitHub stars~2.3k tokensUpdated yesterday
    EducationAuto-check passed
  • Study Companion En

    mingchen666/Reviva

    General study companion for learning, review, exam prep, exam-point analysis, syllabus/exam-outline interpretation, finals speedrun teaching, concept explanation, problem solving, homework coaching…

    236 GitHub stars~5.7k tokensUpdated 16 days ago
    EducationAuto-check passed
  • Learn Law With Rohas

    rohasnagpal/legal-ai-skills

    Acts as an interactive legal tutor for learning a law, legal subject, doctrine, judgment, procedure, or legal concept.

    174 GitHub stars~2.5k tokensUpdated 6 days ago
    EducationAuto-check passed
  • Learning Tutoring

    aipoch/medical-research-skills

    Learning tutoring planning and content production skill for creating study plans, generating exercises, writing answer explanations, and providing review/adjustment guidance; triggered by requests…

    2k GitHub stars~2.8k tokensUpdated 20 days ago
    EducationAuto-check passed

More from XiaomiMiMo/MiMo-Code

All 22 skills in this repo
  • Paper Research on arXiv

    XiaomiMiMo/MiMo-Code

    Searches arXiv, fetches metadata, generates BibTeX, downloads PDFs and finds citations and related papers using a bundled Python script.

    14k GitHub starsUsed in 1 repo~1.5k tokens
    Auto-check passed
  • Agent Skill Creator

    XiaomiMiMo/MiMo-Code

    Interactive guide for creating, reviewing and fixing agent skills (SKILL.md folders), covering structure, frontmatter rules, trigger phrases and validation before sharing.

    14k GitHub stars~1.9k tokensUpdated 4 days ago
    Auto-check passed
  • DOCX Toolkit

    XiaomiMiMo/MiMo-Code

    Produces, edits and reads Microsoft Word files through python-docx and lxml, with a decision table for picking the lightest workflow for a given task.

    14k GitHub stars~2.4k tokensUpdated 4 days ago
    Auto-check passed
  • Drive MiMo Code

    XiaomiMiMo/MiMo-Code

    Lets one MiMoCode process drive another, headless with JSON events or interactively through tmux, to test behavior and visual regressions with parseable evidence.

    14k GitHub stars~3.9k tokensUpdated 4 days ago
    Auto-check passed
  • PDF Toolkit

    XiaomiMiMo/MiMo-Code

    Reads, transforms, composes and fills PDFs with Python scripts for extraction, merging, watermarking, encryption, OCR and form filling.

    14k GitHub stars~1.7k tokensUpdated 4 days ago
    Auto-check passed
  • XLSX Spreadsheet Toolkit

    XiaomiMiMo/MiMo-Code

    Builds, edits, cleans, recalculates and reads Excel workbooks and CSV files with openpyxl and pandas, plus LibreOffice for recalculation and PDF export.

    14k GitHub stars~2.9k tokensUpdated 4 days ago
    Auto-check passed

Categories

Questions about Interactive Course Builder

What does Interactive Course Builder do?

Turns a PDF, paper, document, URL or topic into a chapter-by-chapter course with exercises, feedback and review, and saves progress so you can resume later. The agent builds an adaptive course that alternates explanation, retrieval, application, feedback and review. It shows a course map and then teaches one chapter or lesson at a time, in your language, without revealing exercise answers before you try them.

When should I use Interactive Course Builder?

Interactive Course Builder fits situations like: studying a PDF, paper or book chapter step by step; getting a tutorial or curriculum from source material, with quizzes and hints; continuing a previous course from a saved state file or pasted state block; learning a topic from scratch with a structured course map.

How do I install Interactive Course Builder in Claude Code?

Run `npx skills add XiaomiMiMo/MiMo-Code --skill learn-everything -a claude-code`. Or copy the skill folder (packages/cli/src/skill/builtin/.bundle/learn-everything in XiaomiMiMo/MiMo-Code) into .claude/skills/learn-everything in your project. Claude Code loads it when a task matches its description.

How do I install Interactive Course Builder in Codex?

Run `npx skills add XiaomiMiMo/MiMo-Code --skill learn-everything -a codex`. Or copy the skill folder (packages/cli/src/skill/builtin/.bundle/learn-everything in XiaomiMiMo/MiMo-Code) into .agents/skills/learn-everything in your project. Codex loads it when a task matches its description.

Can I use Interactive Course Builder in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add XiaomiMiMo/MiMo-Code --skill learn-everything -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/learn-everything, .gemini/skills/learn-everything, .github/skills/learn-everything and .opencode/skills/learn-everything in your project.

What does Interactive Course Builder need to run?

SKILL.md names no scripts, command-line tools or credentials: Interactive Course Builder is instructions for the agent only.

Does Interactive Course Builder access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Interactive Course Builder safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Interactive Course Builder use?

Interactive Course Builder is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Interactive Course Builder use?

About 2.9k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1.8k tokens, read only when the agent opens those files.

What are the alternatives to Interactive Course Builder?

Skills that share tags, products or a category with Interactive Course Builder: Claude Certification Tutor (rohitg00/ai-engineering-from-scratch, 65k stars), StudyVault Quiz Tutor (bevibing/tutor-skills, 1.3k stars), Codex Skill Self-Assessment (FlorianBruniaux/claude-code-ultimate-guide, 6.1k stars) and Study Companion En (mingchen666/Reviva, 236 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Interactive Course Builder?

XiaomiMiMo (a GitHub organization) maintains it in XiaomiMiMo/MiMo-Code, which has 13,601 GitHub stars. The repository holds 22 skills in this directory. The repository was last updated on October 3, 2026.

Source: XiaomiMiMo/MiMo-Code on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.