Number Properties content + multi-part Data Insights support (Two-Part & Table Analysis)

[DMDaudio]

What does this PR add or change?

Two related additions to the study platform.

1. Quant → Number Properties content

Fills two previously-empty subtopics with a coherent study unit: 2 lessons + 6 original problem-solving questions (before this, number-properties had a single factors question and no lesson).

• Lessons: Odds/Evens & Signs; Remainders.
• Questions (easy→hard, each with worked explanation + 2 hints):
  • ...odds-evens-signs-0017 (easy, A) — which expression must be even given (x+y) odd.
  • ...odds-evens-signs-0018 (medium, B) — recover the sign of (a) from (abc>0), (bc<0).
  • ...odds-evens-signs-0019 (medium, B) — which expression is always odd (via (n(n+1))).
  • ...remainders-0020 (medium, D) — remainder of (3n+5) mod 7 given (n\equiv4).
  • ...remainders-0021 (hard, C) — remainder of (3^{47}) mod 5 via cyclicity.
  • ...remainders-0022 (hard, D) — least (N>10) with (N\equiv2,(5)) and (N\equiv1,(3)).

2. Multi-part Data Insights support (study-site feature)

Previously the multi-part DI formats had no way to be selected or graded — the site's grader only understood a single answer letter. This adds first-class support:

• New optional parts field — { PartKey: "Opt A|Opt B|Opt C" } — plus a compound answer ("Key=Value; ..."). The Practice view renders one option row per part and grades them together (correct only if every part matches, case-insensitive). Feedback shows the correct option per part, pace, explanation, and hints.
• Single-choice and Data Sufficiency flows were refactored onto a shared feedback tail; their behavior is unchanged.
• Browse view renders each part's options and keyed answer.
• scripts/build.mjs passes parts through to content.json; CONTENT_SCHEMA.md documents the format.
• Seed content (gradeable): a Two-Part Analysis question + lesson (simultaneous-conditions) and a Table Analysis question + lesson (ratios-thresholds).

Rebuilt docs/data/content.json (now 40 questions, 12 lessons) and updated the coverage note in curriculum.md.

Verification: drove the built site in a headless browser — the correct path grades as correct (parts highlighted, explanation shown), a wrong selection grades as wrong (correct answers highlighted), submit gating works, and single-choice/DS questions still grade. No runtime JS errors.

Type of change

• New question(s)
• New lesson(s)
• Fix to existing content (answer / explanation / tag)
• Tooling / docs / schema (study-site multi-part support, build passthrough, schema + curriculum docs)

Author checklist

• This is my own original work — not copied or paraphrased from any real GMAT question or prep material.
• Front-matter follows CONTENT_SCHEMA.md (id matches the filename; topic/subtopic exist in curriculum.md).
• I solved every question myself and confirmed exactly one answer (or correct multi-part combination) is correct.
• Each explanation would help someone who got the question wrong, and names the trap where relevant.
• status is in-review. I did not set my own content to verified.

For reviewers

• I solved the question(s) without reading the keyed answer first and got the same answer.
• Exactly one choice / multi-part combination is correct; distractors are plausible.
• Explanation and tags (section / topic / difficulty) are accurate.
• Multi-part questions render and grade correctly in Practice.
• If approving: bumped status to verified and added my handle to reviewers.

🤖 Generated with Claude Code

https://claude.ai/code/session_01YZoYqoDFwY2nF5BjKJv7qf