feat(iter-11): ship-wheel spinner, progressive hints, risk scorecard & YoY growth levels

[martinl5]

[UI/UX Improvements]

Ship-wheel loading spinner (GameProvider.tsx):

• Replaced the flat shimmer-bar + 🦁 loading screen with an animated SVG maritime ship's helm.
• 8-spoke wheel geometry computed inline via SPOKE_ANGLES.map() (hub r=5.5, spoke tips at r=21, knob circles r=2.8 at r=24, outer rim r=24).
• Rotates at 3 s/revolution via @keyframes spinWheel; no external asset dependency.
• aria-hidden="true" since adjacent text describes the loading state.
• Directly implements the "ship-wheel / compass loading spinners" item from the UI/UX rotation list.

Progressive hint disclosure (SQLPanel.tsx):

• A splitHint() helper splits each level's hint at . boundaries into ≤ 3 progressive chunks (longer hints are grouped into thirds).
• After a first failed submission, a gold 💡 Request hint button appears in place of the full hint text.
• Each click reveals one additional chunk; earlier chunks fade to 70% opacity to de-emphasise them.
• Button label changes to Show more hint (N step(s) remaining) after the first reveal.
• hintChunkIdx state resets to 0 on every level change.
• Changes the cognitive posture from passive delivery to active retrieval — a known pedagogical improvement for self-regulated learning.

---

[Game Design Tweaks]

Level 70 (Expert, difficulty 4) — "Loan Book Risk Scorecard — Weighted Rate & Running Exposure":
Four senior-DS patterns in a single challenge:

1. Weighted-average rate: ROUND(SUM(principal_amount * interest_rate) / SUM(principal_amount), 2) — correct way to aggregate differing rates by volume (wrong: AVG(interest_rate)).
2. Scalar CTE cross-join: a second total CTE feeds the grand total into the final SELECT via CROSS JOIN — the idiomatic pattern when you need a scalar aggregate alongside window functions.
3. Portfolio percentage: ROUND(100.0 * g.total_principal / t.grand_total, 1).
4. Running cumulative exposure: SUM(g.total_principal) OVER (ORDER BY g.risk_grade ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW).

Result: 3 rows (A: $1.88M, 86.5%, cumulative $1.88M | B: $245K, 11.3%, cumulative $2.13M | C: $48K, 2.2%, cumulative $2.17M). Pattern mirrors Basel III RWA dashboards at GIC, JPMorgan, DBS.

Level 71 (Expert, difficulty 4) — "Loan Portfolio Year-over-Year Disbursement Growth":

• CTE yearly: GROUP BY STRFTIME('%Y', start_date) → loan_count, total_disbursed (all statuses — disbursement is a historical fact).
• CTE with_lag: LAG(total_disbursed) OVER (ORDER BY loan_year) AS prev_disbursed.
• Outer: CASE WHEN prev_disbursed IS NULL THEN NULL ELSE ROUND(100.0*(total_disbursed - prev_disbursed)/prev_disbursed, 1) END AS yoy_growth_pct.
• 7 rows (2018–2024): +100% in 2020, −71.4% in 2021, +172.5% in 2022, −87.2% in 2023 — authentic variation that makes the pedagogical point.
• The LAG → CASE WHEN NULL pattern is the canonical SQL idiom for period-over-period growth used in investor reports and loan-book analytics.

---

[Database & Code Optimizations]

• LevelUpModal EPOCH_NEXT_HINT map extended for levels 70 and 71; nextHintKey sentinel array extended to […, 69, 70, 71]; fallback updated ?? 69 → ?? 71.
• MAX_LEVEL in progression.ts auto-derives from levels[levels.length - 1].id = 71 — no other touch points needed.
• No new DB tables, seed rows, or indexes — both levels run against the existing loans table.
• evolution_log.md updated with full iteration-11 entry.

Test Results

Test Files  5 passed (5)
     Tests  395 passed (395)   ← up from 375 in iter-10 (20 new auto-coverage cases for levels 70-71)
  Duration  944ms

npm ci ✅ · tsc --noEmit ✅ · eslint ✅ · vitest run ✅ (395/395)

---

Generated by Claude Code

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
sqlwak	Ready	Preview, Comment	Jun 22, 2026 5:20pm