Technology stack

This page is the technology inventory. The design choices and runtime boundaries are explained in Architecture.

Client

Responsibility	Technology	Role
Application	React 18, TypeScript, Vite	SPA runtime, type safety, and builds
Styling	Tailwind CSS	Shared design tokens and component styling
Routing	React Router	Project and view navigation
State	Zustand, Immer	Project graph, editing history, and runtime state
Canvas	ReactFlow, Dagre	Node graph interaction and automatic layout
Grid	TanStack Virtual	Windowed rendering for large tables
Charts	Recharts	Bar, line, pie, and scatter charts
Reports	TipTap, pdfmake	Rich text editing and PDF generation

Responsibility	Technology	Role
Query engine	DuckDB-WASM	SQL execution inside a Web Worker
Local persistence	IndexedDB, `idb`	Projects, files, reports, and durable sync queues
CSV parsing	PapaParse	CSV import
Workbook handling	SheetJS `xlsx`	Excel import and export
Project archives	JSZip	Self-contained project bundles

Responsibility	Technology	Role
HTTP API	Express	Authentication, project, and file endpoints
Database	MongoDB, Mongoose	Users, project snapshots, and GridFS file storage
Authentication	JWT cookies, bcrypt, Google Auth Library	Password and Google sign-in
API protection	Helmet, CORS, CSRF middleware, rate limiting	Browser and request security controls
Observability	Sentry	Client and server error reporting

The backend is not part of the computation path. DuckDB, project editing, and local persistence continue to work when the API is unavailable.

Continue with Architecture for runtime behavior or Repository structure for the code map.