Generate NEXT-pattern MCQs for a given topic or subject. Creates questions with clinical vignettes, explanations for all options, difficulty tagging, Bloom's taxonomy levels, and textbook references. Outputs as JSON ready for database insertion. Examples: /generate-mcqs "Brachial Plexus" 20, /generate-mcqs AN-MOD-01 50, /generate-mcqs "Pharmacology - Antihypertensives" 30
Create high-quality MCQs aligned with NEXT/NEET PG exam patterns.
$1 = Topic name, topic code, or subject area$2 = Number of questions (default: 20)Invoke the assessment-engine-architect agent with this prompt:
Create $2 NEXT-pattern MCQs for "$1" following these requirements:
- Bloom's distribution: 60% ProblemSolving, 30% Analysis, 10% Recall
- Difficulty distribution: 25% EASY, 40% MEDIUM, 25% HARD, 10% EXPERT
- Each question must have 4 options with exactly 1 correct answer
- Include clinical vignettes for ProblemSolving questions
Output format - Write to prisma/seeds/questions-{topic-slug}.json:
[
{
"type": "MCQ_SINGLE",
"stem": "...",
"clinicalVignette": "...",
"options": [
{"id": "a", "text": "...", "isCorrect": false, "explanation": "..."},
{"id": "b", "text": "...", "isCorrect": true, "explanation": "..."},
{"id": "c", "text": "...", "isCorrect": false, "explanation": "..."},
{"id": "d", "text": "...", "isCorrect": false, "explanation": "..."}
],
"explanation": "Detailed teaching explanation...",
"difficulty": "MEDIUM",
"bloomsLevel": "ProblemSolving",
"examTags": ["NEXT_STEP1", "NEET_PG"],
"textbookRefs": [{"book": "...", "chapter": "...", "edition": "..."}],
"subjectCode": "AN",
"topicCode": "AN-MOD-01-TOP-05"
}
]
Validate each question:
Create a seed script to load these questions into the database.
Report: Print summary of questions created by difficulty, Bloom's level, and exam tag distribution.