gautampartha/smart-sanskriti-

Purpose and Entry Points

Sanskriti is a Next.js heritage-exploration application combining monument discovery, AI assistance, recognition, audio guides, quizzes, gamification, and itinerary planning. The App Router entry is app/page.tsx; global layout and styling begin at app/layout.tsx and app/globals.css. Major user flows are exposed as route pages under app/, including app/explore/page.tsx, app/recognition/page.tsx, and app/dashboard/page.tsx.

Application Shell and Navigation

Shared client UI is organized around reusable shells, navigation, theming, and responsive/mobile presentation. components/app-shell.tsx, components/app-wrapper.tsx, and components/top-nav.tsx establish the main application frame; components/mobile/bottom-nav.tsx provides a mobile-specific navigation boundary. components/auth-guard.tsx protects authenticated experiences.

Heritage Content and Monument Experience

Monument data is handled locally through lib/monuments.ts and lib/monumentStore.ts, then rendered in discovery and detail routes such as app/explore/page.tsx and app/monument/[id]/page.tsx. Recognition-result UI includes monument detail tabs, panorama presentation, and upload handling in components/recognition/monument-detail-tabs.tsx, components/recognition/panorama-viewer.tsx, and components/recognition/upload-zone.tsx. Static audio and time-travel assets are served from public/audio/ and public/time_travel/.

AI Server Boundary

Browser-facing AI features call Next.js server routes, keeping NVIDIA credentials server-side. app/api/chat/route.ts proxies heritage chat, while app/api/recognize/route.ts handles monument image recognition and app/api/transcribe/route.ts handles speech transcription. NVIDIA integration is centralized in lib/nvidia.ts; NVIDIA NIM models and NVIDIA_API_KEY are environment-provided. The related client pages are app/chat/page.tsx, app/recognize/page.tsx, and app/scan/page.tsx.

Personalized Planning and Learning

Itinerary generation is exposed through app/api/itinerary/route.ts and consumed by app/itinerary/page.tsx; planning/domain logic resides in lib/itinerary.ts. Quiz requests flow through app/api/quiz/route.ts, with question content in lib/quizQuestions.ts and presentation in app/quiz/page.tsx. These features feed the broader engagement surface, including app/achievements/page.tsx, app/badges/page.tsx, and app/leaderboard/page.tsx.

Identity, Persistence, and Progression

Authentication and user state are separated into client/context layers: lib/authClient.ts, lib/authContext.tsx, and hooks/useAuth.ts. Supabase is the apparent backend persistence/auth provider, initialized through lib/supabase.ts and lib/supabase/client.ts; XP schema is defined in supabase/migrations/0001_xp_system.sql. Progression logic is represented by lib/xp.ts, while local/offline profile behavior is supported by lib/localProfile.ts.

Runtime and Deployment

This is a TypeScript Next.js runtime configured by next.config.ts, with Tailwind styling configured in tailwind.config.ts. It is deployable to Vercel via vercel.json; deployment must supply NVIDIA environment variables and, where enabled, Supabase credentials. components/BackendPrewarmer.tsx suggests proactive warming of backend/API availability from the client shell.