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}.

Flashcards — Anki-Style Card Generation

Generates study flashcards from topics the student has already learned. Cards are prioritized by weakness, misconceptions, and spaced repetition schedule. Exports in Markdown and Anki-compatible CSV.

Invocation

/claude-teacher:flashcards — generate cards for all studied topics, prioritized by need /claude-teacher:flashcards [topic] — generate cards for a specific topic

Process

1. Load Context

1. Read <education-db>/dashboard.json — get all topic statuses
2. Read <education-db>/student.json — adapt card style to learning preferences
3. If a topic was given, read <education-db>/topics/<topic>.json — check depth, misconceptions, quiz history
4. If no topic given, collect all topics from dashboard and read each topic file
5. NEVER generate flashcards for topics not in the DB — the student hasn't learned them yet. If no studied topics exist: "Nothing to make flashcards for yet — let's learn something first!"

2. Prioritize Topics for Card Generation

Generate cards in this order (highest priority first):

1. Weak topics — these need the most reinforcement
2. Topics with unresolved misconceptions — targeted cards to correct misunderstandings
3. Topics with next_review <= today — due for spaced repetition
4. Topics at surface depth — need more exposure to reach working
5. Learned topics — general reinforcement
6. Solid topics — fewer cards, focused on edge cases and connections

3. Generate Cards

For each topic, generate 5-10 flashcards using a mix of card types. Use WebSearch to verify every fact before putting it on a card — flashcards with wrong information are worse than no flashcards.

Card Types

Definition — core concept recall

Front: What is [concept]?
Back: [Concise definition]. [Why it matters in one sentence].

Concept Application — apply knowledge to a scenario

Front: You need to [scenario]. Which [concept/tool/approach] would you use and why?
Back: [Answer with reasoning]. [Key tradeoff or consideration].

"What happens if..." — trace behavior and consequences

Front: What happens if [specific situation related to topic]?
Back: [Step-by-step what occurs]. [Common mistake: ...].

Compare/Contrast — distinguish related concepts

Front: What is the difference between [concept A] and [concept B]?
Back: [A]: [key trait]. [B]: [key trait]. Key distinction: [one sentence].

Fill-in-the-Blank — active recall of specifics

Front: In [context], the ___ is responsible for [function].
Back: [answer]. [Brief explanation of why].

Card Rules

• Every card back should be concise — aim for 1-3 sentences, not paragraphs
• If the topic file has unresolved misconceptions, create at least one card that directly addresses each misconception
• Use the student's interests from student.json for analogies in card backs (e.g., if they like gaming, use gaming analogies)
• Avoid the student's dislikes — if they dislike dry definitions, make cards more scenario-based
• Include a source link on cards that cite specific technical facts
• Tag each card with difficulty: [easy], [medium], [hard] based on topic depth and quiz history

4. Export in Both Formats

Markdown Format

Save as docs/flashcards-<topic-slug>.md (project-local) and <education-db>/docs/flashcards-<topic-slug>.md (global):

# Flashcards: [Topic Name]

Generated: [date]
Cards: [count]
Priority: [weak / review-due / reinforcement]

---

### Card 1 [easy] — Definition
**Q:** What is [concept]?
**A:** [Answer]. [Why it matters].

---

### Card 2 [medium] — What happens if...
**Q:** What happens if [scenario]?
**A:** [Answer]. Common mistake: [misconception].

---

(... more cards ...)

---

*Generated from your study progress. Review with `/claude-teacher:quiz-me [topic]` or practice with `/claude-teacher:challenge [topic]`.*

Anki-Compatible CSV

Save as docs/flashcards-<topic-slug>.csv (project-local) and <education-db>/docs/flashcards-<topic-slug>.csv (global):

#separator:Semicolon
#html:false
#columns:Front;Back;Tags
#tags column:3
"What is [concept]?";"[Answer]. [Why it matters].";"[topic-slug] [card-type] [difficulty]"
"What happens if [scenario]?";"[Answer]. Common mistake: [misconception].";"[topic-slug] what-happens-if [difficulty]"

CSV rules:

• UTF-8, semicolon-separated; the #separator:Semicolon header makes this explicit
• All fields double-quoted
• Use the metadata headers shown above for Anki 2.1.54+. Do not add a normal Front;Back;Tags row, which could be imported as a card. For older versions, remove metadata and map the separator/fields/tags manually
• Tags field uses space-separated tags: topic slug, card type, difficulty
• Escape any double quotes inside fields by doubling them ("")
• No HTML — plain text only for maximum compatibility

5. Show Summary

After generating, display:

── Flashcards Generated ──────────────────
  Topic: [topic name]
  Cards: [count] ([easy]/[medium]/[hard])
  Misconception cards: [count]
  Card types: [list of types used]

  Saved to:
    docs/flashcards-[slug].md
    docs/flashcards-[slug].csv
    <education-db>/docs/flashcards-[slug].md
    <education-db>/docs/flashcards-[slug].csv

  Import CSV into Anki: File > Import > select the .csv file
  Preview: Front → Front, Back → Back, column 3 → Tags; no header card
──────────────────────────────────────────

If generating for multiple topics, show a combined summary at the end with totals.

6. Update DB

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

{"time": "[now]", "event": "flashcards", "topics": ["[slug1]", "[slug2]"], "total_cards": 15, "formats": ["markdown", "csv"]}

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

• Update last_session to today

Rules

• Only generate cards for studied topics. Check dashboard.json — only topics with status weak, learned, or solid are valid. Never invent cards for topics the student hasn't seen.
• Verify facts with WebSearch. Every factual claim on a card must be checked. Include source links for technical facts.
• Prioritize weak areas. The whole point is targeted review — weak topics and misconceptions come first.
• Keep cards atomic. One concept per card. If a card needs more than 3 sentences on the back, split it.
• Respect student preferences. Use their learning style, interests, and avoid their dislikes when crafting card language.
• Regeneration is safe. Read existing files before regenerating; preserve student edits or save a new version. Keep existing fronts stable when updating answers so Anki can match notes.
• Create project-local docs/ directory if it doesn't exist before saving.

Integration with Other Skills

• /claude-teacher:quiz-me [topic] — flashcards are for self-study; quizzes are for active testing. Suggest quizzing after reviewing flashcards.
• /claude-teacher:challenge [topic] — for hands-on practice after memorizing concepts via flashcards.
• /claude-teacher:progress — check which topics need flashcards most (weak areas, due reviews).
• /claude-teacher:save-progress — flashcard generation auto-logs to the session, but student can explicitly checkpoint.