Requested topic/task: $ARGUMENTS

Before accessing education data, read ${CLAUDE_PLUGIN_ROOT}/references/education-data.md. Resolve every <education-db> below using that contract. Current session ID: ${CLAUDE_SESSION_ID}.

Challenge — Mini-Tasks

Generates a short, focused exercise on the current topic. Adapts to the learning type. Results update the global education DB.

Invocation

/claude-teacher:challenge — picks topic from current context, weak areas, or topics at surface depth that need to reach working /claude-teacher:challenge [topic] — specific topic

Process

1. Read <education-db>/dashboard.json — check topic statuses
2. Read <education-db>/student.json — adapt to learning style and level
3. If a topic was given, read <education-db>/topics/<topic>.json — check depth, misconceptions
4. If no topic given, pick using priority: topics at surface depth (need hands-on to reach working) → weak topics → learned topics
5. NEVER generate challenges on topics not in the DB — the student hasn't learned them yet. If no covered topics exist: "Nothing to challenge on yet — let's learn something first!"
6. Read .claude/claude-teacher.json for learning type (project, subject, exam-prep, or mixed); use CLAUDE.md only as a legacy fallback
7. If topic has unresolved misconceptions, design the challenge to expose/address them
8. Generate ONE challenge appropriate to the type
9. Wait for the student's answer
10. Evaluate, explain, update DB

Challenge Types by Learning Type

Project (code challenges)

CHALLENGE: Write a function that...

  Difficulty: [easy / medium / hard]
  Time: ~5-10 minutes
  Targeting: [depth promotion / misconception / reinforcement]
  Hint available: yes (ask if stuck)

  [Clear problem statement]
  [Expected input/output if applicable]

Subject / mixed (scenario challenges)

CHALLENGE: You are designing...

  Difficulty: [easy / medium / hard]
  Time: ~5 minutes
  Targeting: [depth promotion / misconception / reinforcement]

  [Real-world scenario description]
  [Question: what would you do / choose / configure?]

Exam prep (conceptual or timed challenges)

CHALLENGE: Explain / prove / trace...

  Difficulty: [easy / medium / hard]
  Time: ~5 minutes
  Targeting: [depth promotion / misconception / reinforcement]

  [Problem statement]

Multiple-Choice Challenges — Use AskUserQuestion

When a challenge naturally fits a multiple-choice format (scenario-based "what would you do?" or theory-based "which approach is correct?"), use the AskUserQuestion tool so the student can select with arrow keys. If the tool is unavailable, use plain-text options and wait for an answer.

AskUserQuestion({
  questions: [{
    question: "CHALLENGE (medium): A client reports intermittent timeouts connecting to your service behind a load balancer. Which investigation step would you take FIRST?",
    header: "Scenario",
    multiSelect: false,
    options: [
      { label: "a) Restart the load balancer", description: "Restart the network component" },
      { label: "b) Check load balancer health check logs", description: "Inspect the recorded health checks" },
      { label: "c) Increase the client timeout value", description: "Allow clients to wait longer" },
      { label: "d) Add more backend instances", description: "Increase the number of backends" }
    ]
  }]
})

Rules for AskUserQuestion challenges:

• Provide exactly 4 options
• Randomize the correct answer position
• Write neutral descriptions that do not reveal which choice is correct; explain after the answer
• Keep header at most 12 characters; indicate the challenge type (e.g., "Scenario", "Conceptual")
• If the student selects "Other" (custom input), treat it as an open-ended answer and evaluate accordingly

When NOT to use AskUserQuestion:

• Code challenges (project type) — the student writes actual code, so use plain text
• Open-ended conceptual challenges ("Explain in your own words...") — the student types freely
• Challenges where the answer requires writing, drawing, or multi-step work

Rules

• ONE challenge per invocation. Keep it focused.
• Always state difficulty and estimated time.
• Show what the challenge is targeting (depth promotion, misconception, reinforcement).
• Never give the answer immediately. Wait for their attempt.
• Offer hints on request, not proactively.
• After completing a challenge, suggest: "Want another? Or continue learning?"

Evaluation & DB Update

After the student submits their answer:

1. Evaluate and explain (use WebSearch to verify your evaluation — never hallucinate):

• If correct: acknowledge, explain why it works, highlight good thinking
• If partially correct: credit what's right, explain what's missing
• If wrong: record misconception, explain with source links, offer a simpler version

2. Update topic file <education-db>/topics/<slug>.json:

• If correct and topic was surface depth → promote to working
• If correct with good explanation and topic was working → promote to deep
• If wrong → add misconception, consider demoting status
• Update last_reviewed and recalculate next_review

3. Update dashboard <education-db>/dashboard.json:

• Update topic entry with new status/depth

4. Append to session log <education-db>/sessions/[date].jsonl:

{"time": "[now]", "event": "challenge", "topic": "[slug]", "passed": true, "depth_before": "surface", "depth_after": "working"}

5. Update project-local memory/knowledge_gaps.md if it exists.

Integration with Other Skills

• /claude-teacher:quiz-me — challenges are more hands-on than quizzes. Use challenges to promote depth, quizzes to test breadth.
• /claude-teacher:ascii — if the challenge involves a visual concept, offer to illustrate after evaluation.
• /claude-teacher:progress — student can check how challenges affected their dashboard.