Agent skill

Learn

by kirilxd in kirilxd/claude-tutor

A skill your agent uses when user wants to learn a topic or create a study plan.

MITAuto-check passedEducation

Install Learn

skills CLI
$ npx skills add kirilxd/claude-tutor --skill learn -a claude-code

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

GitHub CLI
$ gh skill install kirilxd/claude-tutor learn --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/kirilxd/claude-tutor.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/learn .claude/skills/learn && 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
GitHub stars
134
Token cost
~3.6k tokens
SKILL.md length
1,681 words
Files
1
Skills in repo
5
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when user wants to learn a topic or create a study plan.

  • Works in 7 steps: Load or Create Learner Profile → Clarify Scope → 5: Diagnostic Assessment (non-beginners… → …
  • User wants to learn a topic
  • SKILL.md covers Overview, File Storage Rules — EXACT…, Process and Topic Slug Convention
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Learn is an agent skill from kirilxd/claude-tutor. Use when user wants to learn a topic or create a study plan. Triggers on "teach me", "I want to learn", "explain X to me", "study", "help me understand", "where do I start with", "how do I get into", or any request to understand a subject in depth. Covers both technical topics (programming, system design, DevOps) and general knowledge (history, science, languages, music theory). Also use when someone asks for a "learning path", "roadmap", or "curriculum" for a topic — even if they don't explicitly say "learn".

Its SKILL.md is about 3.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Education, covering Curriculum and course design and Tutoring and explanations. The repository describes itself as: Turn Claude Code into your personal tutor — personalized learning plans, adaptive quizzes, SM-2 spaced repetition, and a web dashboard. Works with any topic. The licence is MIT.

When your agent uses it

  • User wants to learn a topic
  • Create a study plan
  • I want to learn
  • Explain X to me

Example prompts

  • “teach me”
  • “I want to learn”
  • “explain X to me”
  • “/learn”

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Load or Create Learner Profile
  2. Clarify Scope
  3. 5: Diagnostic Assessment (non-beginners only)
  4. Research
  5. Generate Learning Plan
  6. Save
  7. Teach (optional)

What it can do on your machine

Read from SKILL.md and the folder at commit ce19469. 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 (its code samples are json).

    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

Learn loads about 3.6k tokens when it runs. Until then it costs about 130 tokens; SKILL.md has 1,681 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~130
When it runs · the whole SKILL.md, loaded when a task matches
~3.6k

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 kirilxd/claude-tutor at commit ce19469, republished under its MIT licence (© kirilxd). 1,681 words, ~3,587 tokens.

Download SKILL.mdSave it as .claude/skills/learn/SKILL.md (or your agent's skills folder).
name
learn
description
Use when user wants to learn a topic or create a study plan. Triggers on "teach me", "I want to learn", "explain X to me", "study", "help me understand", "where do I start with", "how do I get into", or any request to understand a subject in depth. Covers both technical topics (programming, system design, DevOps) and general knowledge (history, science, languages, music theory). Also use when someone asks for a "learning path", "roadmap", or "curriculum" for a topic — even if they don't explicitly say "learn".

Learn — Topic Research & Learning Plan Generator

Overview

Guide the user from a vague "I want to learn X" into a structured, researched learning plan with curated resources. Ask clarifying questions, research the topic, and produce a module-by-module plan saved to disk.

File Storage Rules — EXACT PATHS (no deviation)

The learning system uses three separate directories. Each directory has ONE purpose:

DirectoryStoresAllowed fields
~/.claude/learning/plans/Learning plans ONLYtopic, slug, created, level, goal, depth, timeCommitment, modules, totalEstimatedTime, diagnostic
~/.claude/learning/progress/Quiz progress ONLYtopic, quizzes, weakAreas, strongAreas, overallScore, spacedRepetition
~/.claude/learning/ (root)index.json + profile.json ONLYtopics (index), learningStyle/background/createdTopics (profile)

CORRECT — saving a learning plan: ~/.claude/learning/plans/dns-2026-03-29.json WRONG — saving a learning plan: ~/.claude/learning/progress/dns.json

CORRECT — saving quiz progress: ~/.claude/learning/progress/dns.json WRONG — saving quiz progress: ~/.claude/learning/plans/dns-2026-03-29.json

CORRECT — saving any learning data: ~/.claude/learning/plans/dns-2026-03-29.json WRONG — saving to project directory: ./learning/plans/dns-2026-03-29.json

Always use the ABSOLUTE path ~/.claude/learning/ — never a relative path like ./learning/. Never add quiz fields (quizzes, weakAreas, strongAreas, spacedRepetition, overallScore) to plan files. Never add plan fields (modules, resources, goal, depth, timeCommitment) to progress files. Verify the path is correct BEFORE writing.

Process

Check for Existing Topic

Before starting a new plan, check if this topic already exists.

Read ~/.claude/learning/index.json and look for a matching topic (fuzzy match same as quiz skill: "k8s" → "kubernetes").

If the topic exists:

  1. Read the plan file and progress file
  2. Show a brief status:
── Resuming: [Topic] ────────────────────
Level: [level] | Modules: [completed]/[total] | Score: [score]%
Last activity: [date]
──────────────────────────────────────────
  1. Determine the next action:
    • If no quizzes taken on any module: "Ready to start learning? I'll teach you Module 1: [title]"
    • If some modules quizzed: "You left off at Module [N]: [title]. Continue?"
    • If all modules quizzed with score >80%: "You've completed this topic! Want to retake quizzes on weak areas, or adjust the plan?"
  2. If user wants to continue, teach the next incomplete module interactively (see Phase 5: Teach)
  3. If user wants to adjust, load the existing plan and let them modify it

If the topic is new: proceed to Phase 0 (profile) and Phase 1 (scope).

Phase 0: Load or Create Learner Profile

Read ~/.claude/learning/profile.json if it exists.

If profile exists:

  • Greet the user by acknowledging their learning style and background
  • Skip questions about learning style and background in Phase 1 — you already know
  • Still ask topic-specific questions (level, goal, focus areas)

If no profile exists (first time):

  • During Phase 1, include the learning style question (see Phase 1 examples)
  • After Phase 1, save the profile as JSON to ~/.claude/learning/profile.json (create directory if needed):
    • learningStyle: the user's choice (e.g., "hands-on")
    • background: extracted from level + goal answers
    • createdTopics: array with the current topic slug

If profile exists and this is a new topic:

  • Append the topic slug to createdTopics array
Phase 1: Clarify Scope

ALWAYS use the AskUserQuestion tool when asking the user questions, in any context. If you have too many questions for the tool, split them up into multiple calls.

Ask 2-5 clarifying questions, skipping any already answered by the user's initial message.

Rules:

  • Ask ONE question per message — wait for the answer before asking the next
  • Use multiple choice with clear options and descriptions for every question
  • Only one question per message — never combine questions
  • Ask at least 2 (level + goal) — cap at 5
  • If a learner profile exists, skip the learning style question

Question 1 — Current level:

Ask "What's your current experience with [topic]?" with options:

  • No experience — Complete beginner, starting from scratch
  • Some basics — Familiar with core ideas but not hands-on
  • Intermediate — Working knowledge, want to go deeper
  • Advanced — Strong foundation, want expert-level depth

Question 2 — Learning goal:

Ask "What do you want to be able to do after learning [topic]?" with 3 topic-specific goal options. Each option needs a label and a description.

Question 3 — Depth (optional):

Ask "How deep do you want to go?" with options:

  • High-level overview — Understand the big picture and key concepts
  • Working knowledge — Enough to use it confidently day-to-day
  • Deep expertise — Thorough understanding including edge cases

Question 4 — Focus areas (optional):

Ask which areas to focus on. Allow selecting multiple from 2-4 topic-specific subtopics, each with a label and description.

Question 5 — Time commitment (optional):

Ask "How much time do you want to invest?" with options:

  • A few hours — Quick introduction to the basics
  • A weekend — Solid foundation with practice
  • A week — In-depth study with projects
  • Ongoing study — Long-term learning commitment

If no learner profile exists, also ask "How do you prefer to learn?" with options:

  • Reading docs & articles — Text-based, self-paced learning
  • Watching videos — Visual explanations and walkthroughs
  • Hands-on projects — Learn by building and experimenting
  • Theory first — Understand principles, then apply them
Phase 1.5: Diagnostic Assessment (non-beginners only)

Skip this phase if the user said they have "no experience" or are a "complete beginner."

For users with some background, generate a 5-question diagnostic quiz to calibrate their actual level. This prevents wasting time on material they already know.

How to run the diagnostic:

Generate 5 multiple choice questions testing foundational concepts at increasing difficulty. Ask ONE question per message — wait for the answer before the next.

  • Questions 1-2: Basic terminology and concepts (should be easy for anyone with "some basics")
  • Questions 3-4: Intermediate understanding (application of concepts)
  • Question 5: Advanced concept (only experts get this right)

Each question should have 4 options with descriptions. Label them "Diagnostic Q1/5" through "Q5/5".

After the diagnostic, note which level the user actually tested at:

  • 0-1 correct: Suggest starting from beginner despite their self-assessment
  • 2-3 correct: Confirmed intermediate — can skip introductory modules
  • 4-5 correct: Confirmed advanced — compress fundamentals, focus on depth

Store the diagnostic results in the plan JSON by adding a diagnostic field:

json
{
  "diagnostic": {
    "taken": true,
    "score": 3,
    "total": 5,
    "calibratedLevel": "intermediate",
    "skipModules": []
  }
}

Use these results in Phase 3 (Generate Learning Plan) to mark introductory modules as "skippable" or compress them into a quick review.

Phase 2: Research

Perform 3-5 web searches with varied queries:

  1. "[topic] learning roadmap [year]" — find structured learning paths
  2. "best [topic] tutorial for [level]" — find recommended resources
  3. "[topic] official documentation" — find authoritative sources
  4. "[topic] [specific subtopic] guide" — drill into focus areas
  5. "[topic] common mistakes beginners" — anticipate pitfalls

For each search:

  • Use the WebSearch tool
  • Fetch the top 2-3 results with WebFetch
  • Extract: key concepts, recommended order, good resources, common learning paths

Synthesize the research into a coherent structure. Don't just list links — understand what the community recommends and why.

Show full SKILL.md (648 more words)Show less
Phase 3: Generate Learning Plan

Create a structured plan with modules. Present it to the user in readable markdown:

## Learning Plan: [Topic]

**Level:** [beginner/intermediate/advanced]
**Goal:** [user's goal]
**Estimated time:** [total]

### Module 1: [Title]
**Time:** [estimate]
**Objectives:**
- [what the user will understand/be able to do]

**Key concepts:** [list]

**Resources:**
- [resource with link and type]

### Module 2: [Title]
...

Rules:

  • 3-8 modules depending on topic depth
  • Each module should be completable in one sitting
  • Order modules from foundational to advanced
  • Include a mix of resource types per module (docs, videos, tutorials, books)
  • Note free vs paid resources
  • Every module MUST have keyConcepts with at least 2 concept strings — the quiz system uses these to generate questions. A module with empty keyConcepts is broken.
  • If a diagnostic was taken, mark modules covering concepts the user already knows as "Review (optional)" instead of required

Ask the user: "Does this plan look good? Want to adjust anything — add, remove, or reorder modules?"

Phase 4: Save

After user approves the plan:

  1. Construct path: ~/.claude/learning/plans/{topic-slug}-{YYYY-MM-DD}.json
  2. Verify path contains /plans/ — NOT /progress/
  3. Create directory ~/.claude/learning/plans/ if it doesn't exist (use Bash: mkdir -p)
  4. Save plan as JSON to that path — use ONLY the fields listed below:

The JSON format:

json
{
  "topic": "Topic Name",
  "slug": "topic-name",
  "created": "YYYY-MM-DD",
  "level": "beginner|intermediate|advanced",
  "goal": "user's stated goal",
  "depth": "overview|working-knowledge|deep",
  "timeCommitment": "a few hours|a weekend|a week|ongoing",
  "modules": [
    {
      "id": 1,
      "title": "Module Title",
      "objectives": ["objective 1", "objective 2"],
      "keyConcepts": ["concept1", "concept2"],
      "estimatedTime": "2 hours",
      "resources": [
        {
          "title": "Resource Name",
          "url": "https://...",
          "type": "docs|video|tutorial|book|course",
          "free": true
        }
      ]
    }
  ],
  "totalEstimatedTime": "15 hours"
}
  1. Update index at ~/.claude/learning/index.json:
    • Read the existing file first — do not overwrite other topics
    • If the file doesn't exist, create {"topics":{}}
    • Add/update ONLY this topic's entry (preserve all other topics):
json
{
  "displayName": "Topic Name",
  "planFile": "plans/topic-name-2026-03-01.json",
  "progressFile": "progress/topic-name.json",
  "created": "2026-03-01",
  "lastActivity": "2026-03-01",
  "level": "beginner",
  "modulesCompleted": 0,
  "modulesTotal": 5,
  "quizzesTaken": 0,
  "overallScore": null
}
  1. Update profile at ~/.claude/learning/profile.json:

    • Read existing profile (or create {} if missing)
    • Set learningStyle if determined in Phase 1
    • Set background from the level
    • Append the topic slug to createdTopics array (if not already present)
  2. Confirm to the user: "Learning plan saved! You can now use /quiz [topic] to test your knowledge, or /resources [topic] to see all resources."

Phase 5: Teach (optional)

After saving the plan (or when resuming an existing topic), offer to teach the next incomplete module.

Ask: "Ready to start Module [N]: [title]? I can walk you through it now, or you can study on your own and come back for a quiz."

If the user wants to be taught:

Teach the module interactively. The goal is genuine understanding, not information dumping.

Teaching structure for each module:

  1. Connect to prior knowledge — "In the last module you learned [X]. [Topic] builds on that by..." If this is Module 1 or you don't know the user's background, connect to general knowledge or the user's stated goals.

  2. Explain core concepts — Cover each key concept from the module's keyConcepts list. For each concept:

    • Give a clear, concise explanation (2-3 paragraphs max)
    • Use an analogy that connects to something the user likely knows
    • Provide a concrete example
    • If the learner profile says "hands-on", emphasize examples and exercises
    • If "theory-first", lead with principles before examples
  3. Check understanding — After every 2-3 concepts, ask a quick multiple choice comprehension question. This isn't a quiz — it's a "does this make sense?" check. If the user gets it wrong, re-explain differently before moving on.

  4. Reference resources — Point to specific resources from the module: "For a deeper dive on this, check out [Resource Title] — it covers [specific aspect]."

  5. Summarize — At the end of the module, recap the 3-5 key takeaways.

  6. Bridge to next — "Next up is Module [N+1]: [title], which builds on [concept] you just learned."

  7. Suggest quiz — "Want to test what you've learned? Run /quiz [topic] (and mention the module you just covered) to check your understanding."

Teaching guidelines:

  • Keep explanations conversational, not textbook-like
  • Never go more than 3-4 paragraphs without an interaction point
  • If the user seems to be struggling (wrong answers on comprehension checks), slow down and add more examples
  • If the user is breezing through, pick up the pace and add depth
  • Use the learner profile's learning style to adapt your approach

If the user declines teaching:

  • Suggest: "No problem! Study at your own pace using the resources in the plan. When you're ready, run /quiz [topic] to test your knowledge."

Topic Slug Convention

Convert topic to kebab-case for filenames: "Kubernetes Networking" → "kubernetes-networking", "Spanish Grammar" → "spanish-grammar". Use lowercase, replace spaces with hyphens, remove special characters.

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

Files

Just SKILL.md in skills/learn of kirilxd/claude-tutor.

Open the folder on GitHubat commit ce19469

Compare with similar skills

Learn 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.

Learn compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Learn this skillkirilxd/claude-tutor134—~3.6kAutomated safety check: PassMIT
Codex Skill Self-AssessmentFlorianBruniaux/claude-code-ultimate-guide6.1k—~2.3kAutomated safety check: PassCC-BY-SA-4.0
Study Companion Enmingchen666/Reviva237—~5.7kAutomated safety check: PassNone
Agent Skills Learning Path Tutorrohitg00/ai-engineering-from-scratch66k—~1.9kAutomated safety check: PassMIT
Learn Law With Rohasrohasnagpal/legal-ai-skills175—~2.5kAutomated safety check: PassMIT
Interactive Course BuilderXiaomiMiMo/MiMo-Code14k—~2.9kAutomated safety check: PassMIT

Similar skills

  • 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…

    237 GitHub stars~5.7k tokensUpdated 17 days ago
    EducationAuto-check passed
  • Agent Skills Learning Path Tutor

    rohitg00/ai-engineering-from-scratch

    Interactive tutor that teaches one lesson per invocation on creating, discovering, invoking, securing, evaluating, packaging and porting Agent Skills, logging progress to a file.

    66k GitHub stars~1.9k tokensUpdated yesterday
    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.

    175 GitHub stars~2.5k tokensUpdated 7 days ago
    EducationAuto-check passed
  • Interactive Course Builder

    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.

    14k GitHub stars~2.9k tokensUpdated 5 days ago
    EducationAuto-check passed
  • Personal Mba Generator

    aiskilloftheweek/claude-ai-skill-of-the-week

    Creates a comprehensive, structured learning plan to help the user master any skill quickly and effectively.

    146 GitHub stars~1k tokensUpdated 4 mo ago
    EducationAuto-check passed

More from kirilxd/claude-tutor

  • Quiz

    kirilxd/claude-tutor

    A skill your agent uses when user wants to be tested or quizzed on any topic.

    134 GitHub stars~2.7k tokensUpdated 2 mo ago
    Auto-check passed
  • Resources

    kirilxd/claude-tutor

    A skill your agent uses when user asks for learning resources, recommendations, or materials on any topic.

    134 GitHub stars~724 tokensUpdated 2 mo ago
    Auto-check passed
  • Review

    kirilxd/claude-tutor

    A skill your agent uses when user asks about their learning progress or wants study guidance.

    134 GitHub stars~1.2k tokensUpdated 2 mo ago
    Auto-check passed
  • Dashboard

    kirilxd/claude-tutor

    Use only when user explicitly asks to "open dashboard", "show dashboard", "launch dashboard", or "learning UI".

    134 GitHub stars~244 tokensUpdated 2 mo ago
    Auto-check passed

Categories

Questions about Learn

What does Learn do?

A skill your agent uses when user wants to learn a topic or create a study plan. Learn is an agent skill from kirilxd/claude-tutor. Use when user wants to learn a topic or create a study plan.

When should I use Learn?

Learn fits situations like: user wants to learn a topic; create a study plan; I want to learn; explain X to me.

How do I install Learn in Claude Code?

Run `npx skills add kirilxd/claude-tutor --skill learn -a claude-code`. Or copy the skill folder (skills/learn in kirilxd/claude-tutor) into .claude/skills/learn in your project. Claude Code loads it when a task matches its description.

How do I install Learn in Codex?

Run `npx skills add kirilxd/claude-tutor --skill learn -a codex`. Or copy the skill folder (skills/learn in kirilxd/claude-tutor) into .agents/skills/learn in your project. Codex loads it when a task matches its description.

Can I use Learn 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 kirilxd/claude-tutor --skill learn -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, .gemini/skills/learn, .github/skills/learn and .opencode/skills/learn in your project.

What does Learn need to run?

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

Does Learn 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 Learn 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 Learn use?

Learn 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 Learn use?

About 3.6k tokens (SKILL.md is roughly 14k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Learn?

Skills that share tags, products or a category with Learn: Codex Skill Self-Assessment (FlorianBruniaux/claude-code-ultimate-guide, 6.1k stars), Study Companion En (mingchen666/Reviva, 237 stars), Agent Skills Learning Path Tutor (rohitg00/ai-engineering-from-scratch, 66k stars) and Learn Law With Rohas (rohasnagpal/legal-ai-skills, 175 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Learn?

kirilxd (a GitHub user) maintains it in kirilxd/claude-tutor, which has 134 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on July 20, 2026.

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