Underground 2 Revived
Play Need for Speed: Underground 2 (2004) in Chrome and Firefox with WebGPU graphics, Web Audio, up to 120 fps widescreen, controller support, free roam and Online races with friends. Fan project, not
바인더·상자·덱별로 매직 카드를 관리하고, 보유 복사본을 찾으며, 덱리스트를 확인하고 친구들의 교환 상황을 볼 수 있습니다
Track your Magic cards by binder, box, and deck. Find available copies, check decklists, and see what friends have to trade.
게임TypeScript · PLpgSQLGitHub @atalpak온라인운영자 미확인
써 보셨어요? 한 번만 눌러 주세요. 로그인은 필요 없어요.
아직 감지된 업데이트가 없습니다. GitHub 릴리스가 나오면 여기에 보입니다.
개인 계정의 공개 저장소 소유자는 GitHub로 로그인해 운영자를 확인하고 소개를 직접 관리할 수 있습니다. 관리자 승인은 필요하지 않습니다.
로그인한 뒤 확인 파일을 만들어 이 저장소의 기본 브랜치에 30분 안에 커밋해 주세요. 다른 브랜치나 포크의 파일은 인정하지 않습니다. 파일을 올린 뒤 아래에서 확인하면 관리자 승인 없이 관리할 수 있습니다. 자격을 유지하려면 파일을 남겨 두세요.
로그인이 필요하면 로그인 화면으로 이동합니다. 로그인 후 이곳으로 돌아와 원하는 확인 버튼을 다시 눌러 주세요.
프로젝트 폴더에서 /nomorevibe verify를 실행하면 소유권을 확인하고 소개를 직접 관리할 수 있습니다.
1분 만에 확인하기README 전체를 불러오는 중입니다. 아래는 저장된 발췌입니다.
A Magic: The Gathering collection manager that tracks where each copy physically
lives — which binder, which box, which deck — so a digital inventory mirrors
reality, and answers "where is this card, among the people I know?"
No marketplace: nothing can be listed, bought or sold here, and a card links out
to TCGplayer when you want one. Prices are a display-only Scryfall estimate,
reported per finish, saying what could not be priced rather than counting it as
zero.README 전문 보기 ↗A Magic: The Gathering collection manager that tracks where each copy physically lives — which binder, which box, which deck — so a digital inventory mirrors reality, and answers "where is this card, among the people I know?"
No marketplace: nothing can be listed, bought or sold here, and a card links out to TCGplayer when you want one. Prices are a display-only Scryfall estimate, reported per finish, saying what could not be priced rather than counting it as zero.
Status: a working product for a single playgroup, used only by its author so far. Collection and location management, peer-to-peer trading with counter-offers and an atomic transfer on acceptance, decks-as-locations, friends and public profiles, CSV / decklist import, a want list matched against friends' trade binders, a deck playtester, Scryfall-syntax card search, and an Expo mobile app with a live card scanner.
Where the details live — this file is only the front door.
CLAUDE.md: the directory map, data model, hard constraints and commands. Read it before changing anything..claude/rules/: area-specific rules (migrations, app router, data access, testing, mobile)..claude/ORGANIZATION.md: how the agent team works.docs/: feature guides, handoff logs and briefs.archive/: pre-rebuild planning, kept as evidence rather than direction.
| Concern | Choice |
|---|---|
| Web | Next.js 16 (App Router), React 19, TypeScript, Tailwind v4 |
| Database / auth | Supabase — Postgres + Supabase Auth, RLS everywhere |
| Card data | Scryfall: bulk export synced into cards; live search through Scryfall's API |
| Hosting | Vercel free tier; CI on GitHub Actions |
| Mobile | Expo + React Native, npm workspaces (apps/mobile, packages/*) |
Free tier is enough. From Project Settings → API, copy the project URL, the anon key, and the service role key.
cp .env.example .env.local
# fill in NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY,
# and SUPABASE_SERVICE_ROLE_KEY
SUPABASE_SERVICE_ROLE_KEY bypasses row-level security. It is used only by
scripts outside the Next build (the Scryfall sync and the mobile catalog
publisher) and must never reach the browser.
Optional: SIGNUP_INVITE_CODE requires an invite code to sign up (server-only);
SCRYFALL_SEARCH_CONTACT is put in the User-Agent Scryfall requires for search
requests. See .env.example.
With the Supabase CLI:
supabase link --project-ref <your-project-ref>
supabase db push
Or paste the files in supabase/migrations/ into the SQL editor in order.
Name suggestions, collection views and add-to-collection all read the local
cards table, so it must be filled first. (Submitted searches on /search run
on Scryfall's API instead, so they need no local data.)
npm run sync:scryfall # ~500MB download, several minutes
npm run sync:scryfall -- --limit 5000 # quick smoke test instead
npm install
npm run dev
npm run dev # dev server
npm run build # production build
npm run lint # eslint
npm run typecheck # tsc --noEmit
npm test # unit tests over the pure logic in src/lib/**
npm run test:db # apply migrations to a throwaway Postgres and assert on them
Before calling a change done: npm run lint && npm run typecheck && npm test.
npm run test:db needs a reachable Postgres (PGHOST/PGPORT/PGUSER, or
PGURL) and Postgres 15 or newer. It stubs auth.users and auth.uid() so the
schema — including RLS isolation between two users — can be exercised without
booting Supabase. CI runs it against a postgres:16 service container.
The full command list, including the mobile workspaces, is in CLAUDE.md.
scripts/sync-scryfall.ts pulls Scryfall's default_cards bulk export and
upserts every printing into cards. It is a scheduled job, not a one-off script:
updated_at against
the last successful run and skips the download. --force overrides.public.scryfall_sync_runs.It runs daily via .github/workflows/scryfall-sync.yml, which needs
NEXT_PUBLIC_SUPABASE_URL and SUPABASE_SERVICE_ROLE_KEY as repository secrets.
The same workflow refreshes the Oracle-card table and publishes the mobile
catalog.
Scryfall is a free service run on donations. Keep the sync at daily or slower —
the bulk data is only regenerated about that often — and keep search traffic
inside the shared limiter (claim_scryfall_slot, migration 48).
Every user-owned table has row-level security, but RLS is the floor, not the
whole filter: since a friend's tradable binder is readable by design, queries
that ask "what do I own" must also filter on the owner explicitly, and removing
one of those filters is a cross-user data leak. Nothing under src/, apps/ or
packages/ may hold a database credential stronger than the signed-in user's.
Trade transfers happen inside the public.accept_trade SECURITY DEFINER
function, never through a loosened policy, and ownership_history rejects
UPDATE and DELETE at the trigger level. The reasoning and the full list of
hard constraints are in CLAUDE.md.