A centralised dashboard that tracks the LeetCode performance of every student
and staff member in the college.
Only the administrator signs in. Students and staff never create accounts —
the admin uploads one spreadsheet, and a scheduled job reads each person's
public LeetCode profile once a day and stores a permanent daily snapshot.
Every report, chart and leaderboard is built from those snapshots.
Runs entirely on free tiers: Supabase Postgres + Vercel Hobby + GitHub Actions.
Contents
Features
Dashboard — 12 KPI tiles (total users, active today, solved today, highest
streak, averages by difficulty, weekly/monthly growth with trend arrows), a
30-day activity chart, a college-wide difficulty donut, top-performer /
most-active / longest-streak highlight cards, and a department comparison.
Users — searchable, sortable, filterable roster with inline add/edit/delete,
per-user "refresh from LeetCode", and Excel/CSV/PDF export honouring the active
filters.
Individual profile — identity card with college and department rank, seven
period counters (today / week / month / year / streak / best streak / contests),
plus daily-activity, cumulative-progress, streak-history and global-ranking
charts.
Daily Report — a dated snapshot of every profile. Pick any past date and the
table reproduces that day exactly, with page totals and print support.
Leaderboard — daily / weekly / monthly / all-time, with a podium for the top
three. Weekly and monthly are summed from snapshots, not guessed.
Split Report — how the college distributes across problems-solved bands
(0, 1–10, 11–25, 26–50, 51–100, 101–200, 201–500, 501–1000, 1000+). Click a band
to list exactly who is in it, and export just that band.
Department & Staff analytics — per-department totals, averages, streaks,
active-today share, weekly volume, difficulty mix, and top / most-active /
inactive callouts. Staff get a dedicated page separate from students.
Bulk import — drag-and-drop .xlsx / .xls / .csv with fuzzy header matching
("Reg No" and "Register Number" both work), row-level validation, and an import
summary listing every skipped row with the reason.
Notifications — generated after each sync: new highest streak, top daily
solver, inactive users, invalid profiles, and username changes.
Settings — sync schedule (with the exact UTC cron expression to paste into
the workflow), inactivity threshold, sync history, and JSON backup/restore.
Plus: dark mode, keyboard search (⌘K), responsive down to mobile, and
the college crest throughout.
Architecture
Administrator (single account)
│
Next.js App Router UI
│
Next.js API routes
│
Prisma ── PostgreSQL
│
┌─────────────────────┴─────────────────────┐
│ │
users (current values) daily_snapshots (history)
│ │
└─────────────────────┬─────────────────────┘
│
GitHub Actions · once per day
│
LeetCode public GraphQL endpoint
Two levels of storage, deliberately:
users holds each person's latest known figures. Every list, filter and
leaderboard reads from here, so those pages need no joins.daily_snapshots holds one immutable row per user per day. Every
historical report, trend chart and weekly/monthly total is derived from here.
todaySolved is computed as the difference between today's cumulative total and
the most recent previous snapshot — LeetCode does not expose "problems solved
today" directly.
Tech stack
| Layer | Choice |
|---|
| Framework | Next.js 15 (App Router) · React 19 · TypeScript |
| Styling | Tailwind CSS v4 + hand-rolled shadcn-style components |
| Charts | Recharts |
| Database | PostgreSQL (Supabase free tier) |
| ORM | Prisma 6 |
| Auth | Auth.js (NextAuth v5), credentials, admin only |
| Spreadsheets | SheetJS (xlsx) |
| PDF export | jsPDF + autoTable |
| Scheduler | GitHub Actions cron |
| Hosting | Vercel Hobby |
No paid service is required at any point.
Quick start
Prerequisites: Node.js 20+ and a PostgreSQL database (local or Supabase).
git clone <your-repo-url> leettrack-ai
cd leettrack-ai
npm install
Create your environment file:
cp .env.example .env
Fill in DATABASE_URL, then generate the two secrets:
openssl rand -base64 32 # → AUTH_SECRET
openssl rand -hex 32 # → CRON_SECRET
Set ADMIN_EMAIL and ADMIN_PASSWORD to the credentials you want, then:
npm run db:deploy # apply migrations
npm run db:seed # create the administrator account
npm run dev
Open http://localhost:3000, sign in, and upload your spreadsheet at
Bulk Import. A ready-made example lives at
docs/sample-import.xlsx.
Then run one sync to populate the statistics:
npm run sync
以上为 README 的开头部分。其余内容请阅读原文。 查看完整 README ↗
