Your GitHub, rated out of 99 — as a cricket card.
Live: https://gitcric.dev
demo2-github.mp4
26-second demo — scouting a profile, format switching, and the card download.
git clone https://github.com/bhabishnu/gitcric.git
cd gitcric/web
npm install
npm run devThen open http://localhost:3000.
A GitHub token is optional but recommended for higher rate limits — create web/.env.local
containing GITHUB_TOKEN=ghp_yourtoken. The token needs no scopes (it reads public data
only) and never reaches the client.
The app runs entirely on prebuilt static data in web/gen/ — no database and
no data pipeline setup is needed to run the web app. (The src/ ETL only matters
for regenerating cricketer data.)
gitcric.dev/<your-username>
That's the whole product. You get a card, a scout report of the signals behind it, and the cricketer you'd be — toggle Test · ODI · T20I · IPL to meet a different twin in each format. Kohli in T20Is doesn't mean Kohli in Tests.
The card also flies your flag, read from your GitHub location, and marks your top language — both pulled off the profile, same as the numbers.
Six signals off a live GitHub profile. No surveys, no self-reporting, nothing you can fill in about yourself.
| Stat | Cricket meaning | What it reads |
|---|---|---|
| BAT | Batting | total stars earned, weighted by your biggest repo |
| POW | Power | contributions in the last year — raw velocity |
| BWL | Bowling | pull requests opened on other people's repos, issues closed |
| ECO | Economy | focus: how concentrated your stars are, how tight your language stack is, how many PRs you actually land vs open |
| FLD | Fielding | code reviews, plus followers |
| IMP | Impact | lifetime contributions and how many years you've been active |
ECO is the one people argue with. It rewards finishing things — a maintainer with one flagship and a high merge ratio reads high; forty half-abandoned repos across eleven languages reads low. That's deliberate, and it's the closest honest analog to a bowler who doesn't leak runs.
Two bands, borrowed from how the cricketers are scored. Your six stats are role-weighted into a peak score that caps out — no matter how good the stats are, that alone won't carry you past the low 80s. Everything above is a greatness band, bought with sustained reach and years on the clock rather than a good twelve months, which is why 99 is effectively unreachable and even 96 is rare.
The point of the cap is the twins. User OVRs are calibrated onto the same 40–99 scale as the real cricketers in the database, so "you're a 94" and "Muralitharan is a 92" are the same sentence. Without that, matching you to a cricketer would be decoration. With it, it's a lookup.
The cricketer side is stricter still: peak caps at 88, the greatness band runs 88→99, and 99 is pinned editorially — Kohli's ODI card is #1, Tendulkar's ODI #2.
The rim and finish on the card change with the number.
bronze under 70 · silver 70–83 · gold 84–90 · immortal 91+
The header line above the card names it differently — ON THE FRINGE, IN THE SQUAD, CAPTAIN'S PICK, GENERATIONAL — because a tier is a material and a selection is a verdict, and they're not the same thing.
1,684 cricketers, 2,698 format cards, built from Cricsheet ball-by-ball data through a SQLite pipeline that lands as static JSON in the web app. Nothing queries a database at request time. Qualification gates keep the pool honest — 8 Tests, 17 ODIs, 25 T20Is, 20 IPL matches, below which you don't get a card at all, which is what stops a two-cap wonder turning up as your twin.
Twins are drawn from cricketers gated in at least two formats, sampled within ±3 OVR of you (±4 up at the top, where the pool thins), seeded on your username so your card never reshuffles.
335 player photos come from Wikimedia Commons under free licences, credited in
web/public/players/CREDITS.md. Where Commons has
nothing usable, the card falls back to a monogram rather than ship something wrong.
Built with Next.js 15 · TypeScript · Tailwind v4 · Barlow Condensed & JetBrains Mono,
plus the SQLite → static-JSON data pipeline in src/.
GitCric is a port of GitFut by @Younesfdj — the same idea, rated as football. The two-band scoring engine here started as his, re-axed from PAC/SHO/PAS to BAT/POW/BWL and recalibrated against a cricket population. GitFut is MIT-licensed; the notice travels with the code in NOTICE.
If you like this, go star his repo — the good idea was his first.
Pipeline docs: docs/pipeline.md · Cricket data © Cricsheet,
CC BY 3.0 · MIT