Resumorah

Resumorah is an AI-assisted CV builder: regional templates, 12 languages and PDF export, built with Next.js, MongoDB and Auth.js.

Brand

• Logo: a rounded R with a four-point spark beside it ("orah" means light; the spark is also the product's AI cue). The mark lives in components/ui/logo.tsx (BRAND_MARK_PATH, BRAND_SPARK_PATH). Use <Logo /> for the lockup and <LogoMark variant="tile" /> for the app-icon tile.
• Wordmark: lowercase resumorah in Syne Bold.
• Colours: Rose #e11d48 → Tangerine #fb923c (the brand gradient). The coral theme in app/tokens.css is the default site theme.
• Icons: app/icon.svg, app/favicon.ico, app/apple-icon.png, public/icons/* and the web manifest in app/manifest.ts. The social preview is app/opengraph-image.png / app/twitter-image.png.

Local development

1. Copy .env.example to .env.local and fill in values (see Environment variables).
2. Install and run:

npm install
npm run dev

Open http://localhost:3000.

Themes, dark mode and languages

The globe button (navbar, auth pages, dashboard sidebar, builder header) opens the preferences menu:

• Languages: English, Español, Français, Deutsch, العربية (right-to-left) and 中文. Strings live in lib/i18n/dictionaries/; en.ts is the source shape and every other locale is type-checked against it. lib/i18n/dictionaries.test.ts checks placeholders and array lengths.
• Themes: Violet, Ocean, Emerald, Rose and Sunset. Brand colours are CSS variables in app/tokens.css.
• Appearance: Light, Dark or System.

Choices are stored in the cvg-locale, cvg-theme and cvg-mode cookies and applied on the server, so pages render without a flash. Without a cookie, the language comes from the browser's Accept-Language. Dark sections and the CV paper use the scheme-light class to keep the light palette. Legal documents and server error messages are English only.

Scripts

Command	Purpose
npm run dev	Development server
npm run build	Production build
npm run start	Run production build locally
npm run lint	ESLint
npm test	Unit tests (tsx --test lib/**/*.test.ts)

Environment variables

All configuration is server-only (this app does not use NEXT_PUBLIC_* for secrets).

Variable	Required	Notes
MONGODB_URI	Yes (runtime)	MongoDB connection string
MONGODB_DB_NAME	Yes (runtime)	Database name (use separate names per environment)
AUTH_SECRET	Yes in production	openssl rand -base64 32
AUTH_URL	Recommended in production	e.g. https://your-app.vercel.app
AUTH_GOOGLE_ID	Optional	Both Google vars required to enable Google sign-in
AUTH_GOOGLE_SECRET	Optional	OAuth client secret
OPENROUTER_API_KEY	Optional	Enables AI features via OpenRouter; disabled when unset
AI_MODEL	Optional	OpenRouter model id; defaults to openai/gpt-4o-mini
OPENROUTER_SITE_URL / OPENROUTER_APP_NAME	Optional	OpenRouter attribution headers (site URL falls back to AUTH_URL)

See .env.example for a safe template to commit. Never commit .env or real credentials.

Production deployment (Vercel)

Prerequisites

• MongoDB Atlas (or compatible) cluster
• Vercel project linked to this repository
• (Optional) Google Cloud OAuth client for Google sign-in
• (Optional) OpenAI API key for AI assist

MongoDB

1. Create a cluster and database user with read/write access.
2. Allow network access from Vercel (Atlas: Network Access ΓåÆ allow 0.0.0.0/0 or Vercel static IPs if you restrict).
3. Copy the connection string into Vercel as MONGODB_URI.
4. Set MONGODB_DB_NAME to a production-specific name (e.g. cv-generator-production).

The app reuses a cached Mongoose connection across serverless invocations (lib/db/connect.ts).

Auth.js on Vercel

1. Set AUTH_SECRET (32+ random characters) for Production (and Preview if you test auth there).
2. Set AUTH_URL to your deployment URL, e.g. https://your-app.vercel.app.
3. trustHost is enabled in auth.config.ts for hosted deployments.

Email/password signup works when MongoDB and AUTH_SECRET are configured.

Google OAuth (optional)

In Google Cloud Console:

1. Create an OAuth 2.0 Web application client.
2. Authorized redirect URI: https://your-app.vercel.app/api/auth/callback/google
   (Add http://localhost:3000/api/auth/callback/google for local dev.)
3. Set AUTH_GOOGLE_ID and AUTH_GOOGLE_SECRET in Vercel.

AI (optional)

Set OPENROUTER_API_KEY (get one at https://openrouter.ai/keys) in .env.local or Vercel — that's the only required step. This enables AI summary generation, work-experience rewriting, skill suggestions and resume import parsing.

Optionally set AI_MODEL to any OpenRouter model id, e.g. anthropic/claude-3.5-haiku or google/gemini-2.0-flash-001.

Deploy on Vercel

1. Import the Git repository in Vercel.
2. Framework preset: Next.js
3. Build command: npm run build (default)
4. Install command: npm install (default)
5. Add environment variables for Production (and Preview as needed).
6. Deploy.

PDF export uses the Node.js runtime (@react-pdf/renderer) on GET /api/cv/[id]/export.

Post-deployment smoke tests

Run through this checklist on the production URL:

1. Homepage loads
2. Signup (email/password) works
3. Google login works (when Google OAuth is configured)
4. Login and logout work
5. Dashboard loads for signed-in users
6. CV creation works
7. CV editing and autosave work
8. Template switching works
9. AI features work (when OPENROUTER_API_KEY is set)
10. PDF export downloads a valid file
11. Unauthorized CV access is rejected (signed out or another userΓÇÖs CV id)
12. Production build succeeded in Vercel (deploy logs)

Architecture notes

• Auth: JWT sessions; dashboard routes protected via proxy.ts (Next.js 16 proxy convention).
• CV access: Server actions and APIs enforce ownership via getCurrentUser and checkCvOwnership.
• PDF export: Server reads saved CV from MongoDB; client sends only the CV id.