Skip to content

FAQ Generator Skill

Summary

This skill generates a comprehensive set of Frequently Asked Questions (FAQs) from course content, learning graphs, and glossary terms to help students understand common questions and prepare content for chatbot integration.

Order

This skill should be executed after the following artifacts exist:

  1. Course description has been finalized
  2. Learning graph has been created
  3. Glossary has been generated
  4. At least 30% of chapter content has been written

Having these prerequisites ensures the FAQ generator has sufficient context to create meaningful, relevant questions.

Inputs

Primary Input Files

  1. Course Description (docs/course-description.md)
  2. Provides course scope and objectives

  3. Quality check: Must contain clear learning outcomes using Bloom's Taxonomy

  4. Learning Graph (docs/learning-graph/03-concept-dependencies.csv)

  5. Identifies key concepts and their relationships

  6. Quality check: Should be a valid DAG with no cycles

  7. Glossary (docs/glossary.md)

  8. Source of terminology for technical questions

  9. Quality check: Minimum 50 terms for comprehensive FAQ

  10. Chapter Content (all docs/**/*.md files)

  11. Provides context for specific topic questions

  12. Quality check: At least 5,000 total words preferred

  13. Existing FAQ (docs/faq.md) - if present

  14. Preserves manually curated questions
  15. Merges with AI-generated questions

Input Quality Metrics (Scale 1-100)

Content Completeness Score: - 90-100: All inputs present with high quality - 70-89: Core inputs present, some content gaps - 50-69: Missing optional inputs or low word count - Below 50: Critical inputs missing

Quality Checks:

  1. Course description completeness (title, audience, prerequisites, outcomes)
  2. Learning graph validity (no cycles, reasonable connectivity)
  3. Glossary size (50+ terms = good, 100+ = excellent)
  4. Content word count (target: 10,000+ words for comprehensive FAQ)
  5. Concept coverage (what % of learning graph concepts have related content?)

User Dialog Triggers: - Score < 60: "Limited content available for FAQ generation. Continue with basic FAQ or wait for more content?" - No glossary: "No glossary found. Generate FAQ anyway (limited technical questions) or create glossary first?" - Low word count: "Only [N] words of content found. FAQ quality may be limited. Proceed?"

Outputs

Generated Files

  1. docs/faq.md - Complete FAQ in categorized format
  2. Header: "# Intelligent Textbooks FAQ"
  3. Categories aligned with learning graph taxonomy
  4. Questions organized from basic to advanced
  5. Answers with links to relevant sections

  6. Format: Level-2 headers for questions, body text for answers

  7. docs/learning-graph/faq-quality-report.md - Quality metrics

  8. Bloom's Taxonomy distribution of questions
  9. Concept coverage analysis (what % of concepts have FAQ entries)
  10. Question difficulty distribution

  11. Suggested additional questions for gaps

  12. docs/learning-graph/faq-chatbot-training.json - Structured data for chatbot

  13. JSON array of question-answer pairs
  14. Metadata: difficulty, concepts, keywords
  15. Cross-references to source content
  16. Ready for RAG system integration

Output Quality Metrics (Scale 1-100)

FAQ Quality Score Components:

  • Coverage (30 points):
  • What % of learning graph concepts addressed?
  • Target: 80%+ coverage = 30 points
  • 60-79% = 20 points

  • Below 60% = 10 points

  • Bloom's Taxonomy Distribution (25 points):

  • Balanced across levels (Remember, Understand, Apply, Analyze, Evaluate, Create)
  • Target distribution: 20% Remember, 30% Understand, 25% Apply, 15% Analyze, 7% Evaluate, 3% Create

  • Deviation from target reduces score

  • Answer Quality (25 points):

  • Answers are complete and accurate
  • Include examples where appropriate (target: 40% with examples)
  • Link to relevant content sections (target: 60%+ linked)

  • Appropriate length (100-300 words average)

  • Organization (20 points):

  • Logical categorization aligned with course structure
  • Progressive difficulty within categories
  • No duplicate questions
  • Clear, searchable question phrasing

Question Categories Generated:

  1. Getting Started (10-15 questions)
  2. Core Concepts (20-30 questions based on learning graph)
  3. Technical Details (15-25 questions from glossary)
  4. Common Challenges (10-15 questions)
  5. Best Practices (10-15 questions)
  6. Advanced Topics (5-10 questions)

Quality Checks Performed:

  1. No duplicate questions (100% unique)
  2. All internal links valid
  3. Bloom's Taxonomy distribution
  4. Reading level appropriate for target audience
  5. Answer completeness (no partial answers)
  6. Technical accuracy (verified against glossary/content)

Success Criteria: - Overall quality score > 75 - Minimum 40 questions generated - At least 60% concept coverage - Balanced Bloom's Taxonomy distribution - All answers include source references - Chatbot JSON validates against schema

Additional Outputs

  1. Navigation Update - Ensures FAQ link exists in mkdocs.yml
  2. docs/learning-graph/faq-coverage-gaps.md - Lists concepts without FAQ coverage for future content creation