Note on Create-level questions: The Create level (designing new artifacts) maps poorly to single-answer multiple-choice format. The skill spec allows rotating the advanced R/C slot — all seven advanced-chapter slots were allocated to Remember. Open-ended Create-level exercises would belong in a separate hands-on lab section, not in the MCQ quiz files.
Answer balance score: Acceptable but skewed. Letter D is underrepresented in the overall set (15% vs. 25% target). Eight chapters fall outside the 2-3-per-letter window — most notably Ch 4 (B=5, D=0), Ch 8 (A=5), Ch 9 (B=6), Ch 13 (C=5), Ch 18 (A=5), Ch 22 (A=5). Each individual question's letter was chosen for distractor plausibility rather than for balance. Future passes may rebalance by re-ordering options on selected questions.
All 22 chapters contain 3,500+ words. The smallest source is Ch 12 (3,555 words); the largest is Ch 20 (8,129 words). Content readiness was excellent across the entire book — no chapter required scaling down to fewer than 10 questions.
Each quiz tests concepts drawn from the chapter's own narrative (named patterns, defined terms, illustrative scenarios, and contrasts) rather than from external trivia.
Distractors are drawn from related concepts within the surrounding chapter material or from commonly confused alternatives (e.g., RDF vs. LPG, lineage vs. provenance, registry vs. catalog, MCP vs. function calling).
Rebalance D-letter answers. A future pass could shuffle option order on ~10 questions in chapters 4, 9, 11, 13, 15 to lift D from 15% toward 25% without changing question content.
Consider a Hands-On Labs section. Create-level cognitive work (schema design, ingestion pipeline architecture, graduated-autonomy rollout plans) belongs in open-ended labs rather than MCQ quizzes — Chapters 13, 15, 16, 19, 21 are the natural homes.
Spot-check the heavy-A chapters (8, 18, 22). Each has 5 of 10 correct answers as letter A. Verify that this isn't a giveaway pattern when students take quizzes sequentially.