feat: refresh auth providers and score precision

[quan0715]

變更目標

• 將登入 provider 設定收斂到單一 provider registry / register_oauth_provider contract
• 收斂 NYCU provider key，遷移舊 nycu-oauth identity 資料
• 補齊 NYCU login showcase 與 QAuth 身份認證擴充文件
• 合併目前分支上的考試監控與分數精度調整

影響範圍

• backend/apps/users/auth, backend/apps/users/views/auth.py, users migrations/tests
• backend/apps/contests score / anticheat 相關服務與 migration
• frontend auth entry screens、contest score display、docs pages

驗證

• Pre-push hook: ESLint 通過（399 warnings, 0 errors）
• Pre-push hook: TypeScript 編譯通過
• Pre-push hook: frontend production build 通過
• Pre-push hook: Docker Compose config 通過
• backend users tests: pytest -q apps/users/tests tests/test_users.py 通過（150 passed）
• migration check: python manage.py makemigrations --check --dry-run 通過
• docs build: npm run build:docs 通過

風險與回滾

• 涉及 auth provider registry、users migration、contest score precision migration；若上版後異常，回滾此 PR 並回復 migration 狀態。

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: c491538085

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Use half-up decimal rounding for scores

When value is a valid score such as 1.005 (which can be produced by redistribution ratios before persistence), binary floating-point makes numeric * 100 fall just below the half, so this returns 1 and formats 1.00; the backend scorer now persists totals with Decimal(..., ROUND_HALF_UP), so admin previews/UI formatting can disagree with saved totals by 0.01. Please use a backend-equivalent decimal/string rounding path here rather than raw Math.round on a float.

Useful? React with 👍 / 👎.