Quiz

Arguments

The user invoked this with: $ARGUMENTS

Parse n from $ARGUMENTS as the number of questions to run (default: 1).

How to run a question

1. Execute the question generator:

python3 $CLAUDE_PLUGIN_ROOT/skills/quiz/scripts/gen_question.py

If the script fails (non-zero exit or malformed output), inform the user the type data is unavailable and suggest running /type-chart first to confirm the data cache is intact.

2. The script outputs JSON with fields: question, answer, mult, explanation

3. Present ONLY the question field to the user. Do NOT reveal the answer or explanation yet.

4. Wait for the user's response.

5. Evaluate their answer against the answer field:

   • For multiplier questions: accept reasonable approximations (e.g. "super effective", "1.6", "SE" all count for a 1.6x answer)
   • For list questions: require all correct types; minor ordering/spelling ok. No partial credit — if they miss any type, mark incorrect but show what they missed in the feedback.
   • Partial credit: if they get the multiplier bucket right (SE/NVE/neutral/immune) but not the exact value, count as correct with a note

6. Give feedback:

   • Correct: confirm + reinforce WHY (use the explanation field)
   • Wrong: give the correct answer + explanation, then connect it to a real in-game scenario if possible

Scoring (multi-question sessions)

Track score across the session. At the end report: X/N correct, flag any category that appeared more than once and was missed.

Question types (weighted)

• Single type matchup (1x weight)
• Dual type defender matchup (2x weight — most relevant in-game)
• List all SE attackers vs a type (1x weight)
• Double weakness on a dual type (1x weight)