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
경제·세력·NPC·역사를 갖춘 살아있는 테이블탑 RPG 마을을 만들고 지속 세계로 시뮬레이션합니다
Generate living tabletop-RPG settlements with economies, factions, NPCs, and history, then simulate them as a persistent world.
게임JavaScript · HTMLGitHub @clausellstokes-lang온라인운영자 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 확인하기A procedural D&D settlement generator that produces coherent, playable locations with integrated economy, factions, NPCs, and active tensions. Designed for immediate table use, combining structured systems with optional AI synthesis for fast, high-quality worldbuilding.
저장소 README에서 가져옴
README 전체를 불러오는 중입니다. 아래는 저장된 발췌입니다.
A constraint-driven D&D settlement simulator (not a table-roller). A config
(tier, terrain, culture, trade access, threat, priority sliders) is pushed through
a deterministic, seeded pipeline that produces an internally-coherent settlement —
economy, factions, institutions, NPCs, stressors, history — rendered as an on-screen
dossier and an exportable PDF, with an optional AI prose layer and a living-world
campaign simulation (world-pulse ticks over a regional map).README 전문 보기 ↗A constraint-driven D&D settlement simulator (not a table-roller). A config (tier, terrain, culture, trade access, threat, priority sliders) is pushed through a deterministic, seeded pipeline that produces an internally-coherent settlement — economy, factions, institutions, NPCs, stressors, history — rendered as an on-screen dossier and an exportable PDF, with an optional AI prose layer and a living-world campaign simulation (world-pulse ticks over a regional map).
Stack: React 19 · Zustand 5 · Vite 7 · JS with JSDoc types · Supabase (auth,
Postgres + RLS, edge functions) · Stripe (credits / subscription / one-shot dossiers)
· Anthropic (AI narrative). No .ts in app code; the engine is pure and headless.
data → generators → presentation), the generation pipeline, state
(Zustand slices), routing, the Supabase backend, the design system, and the gate.
Read this first; it is kept in sync with the code, and its completeness claims
(the "machine-enforced" / "promoted to ERROR" vocabulary) are gate-checked against
live enforcers (tests/docs/enforcement-claims.test.js). Other prose there is
maintained by review, not by a gate.
docs/ corpus, split into
canonical (trust as current) and historical (point-in-time snapshots;
when a historical doc and the code disagree, the code wins).npm install
npm run dev # Vite dev server
npm run check # the full gate (see below) — run before pushing
npm test # Vitest only
npm run test:e2e:performance # production-build browser regression receipt
Node ≥ 22, npm ≥ 10 (.nvmrc pins the major). Copy .env.example → .env and fill
the VITE_* vars for Supabase/Stripe; secrets live in the Supabase/Vercel dashboards,
never in the repo (the anon key is public by design — RLS enforces access).
npm run check runs the whole && chain declared by the check script in
package.json, which is where the stage count lives — this page names the stages
and never counts them, because a second copy of that number is a copy that drifts.
In order: hazard-registry, premortem and implementation-packet validation;
data-key, custom-content-manifest, and migration-ledger validation; edge,
map-bridge, tuning-band, Foundry-module, and MCP-server contract checks; the
full-tree and strict-domain typecheck ratchets; lint; the Vitest suite through its
per-test census; the production build; and built-artifact verification. It runs in CI
on every push/PR and via a husky pre-push hook; production deploys are
fail-closed gated on CI (scripts/vercel-ignore-build.mjs). Edge (Deno) tests run
via npm run check:edge-behavior / the CI deno-tests job. The money/security
coverage floors (npm run test:coverage:floors) run as a separate required CI job,
and the production-build browser performance receipt runs in its own CI job
(npm run test:e2e:performance locally). Neither is inside npm run check, so a
local green check has not exercised them. See
ARCHITECTURE.md → "The gate" and docs/DEPLOY.md.
Solo-developer project. English-only today — copy is centralized behind a t()
indirection seam (src/copy/), but there is no locale-selection layer yet, so treat
"i18n" as a seam, not a shipped capability. Bus factor is one; ARCHITECTURE.md and the
small canonical reading set in docs/README.md exist to lower the cost of a second
contributor. The wider docs corpus contains historical design and execution records;
volume is not the same as current authority.