China EV DB

A database of Chinese-market EVs/ICE/hybrids, cross-referenced against Morocco-market pricing and availability. Browse and compare vehicles by brand, spec, and price; research is AI-assisted (real web search + an LLM, see below).

Stack: Next.js (App Router) + MongoDB/Mongoose, deployed on Vercel.

Setup

npm install
cp .env.example .env.local

Fill in .env.local:

• MONGODB_URI — a MongoDB connection string.
• AI_GATEWAY_API_KEY — for AI-assisted research (spec/brand/pricing lookups). Create one at Vercel AI Gateway, or leave blank and run vercel link && vercel env pull for local OIDC auth instead.
• SEARCH_API_KEY — a free Brave Search API key, used to fetch real search results before any AI call.

npm run dev

Open http://localhost:3000.

Project layout

• app/ — Next.js pages, API routes, and client components.
• models/ — Mongoose schemas (Brand, Model, Powertrain, MoroccoListing).
• lib/ — shared logic: the AI research pipeline (aiProvider.ts, webSearch.ts, groundedResearch.ts, techSpecResearch.ts, brandResearch.ts, modelDiscovery.ts), data scrapers, and display helpers.
• scripts/ — CLI tools for batch research/import/price-sync (run via npm run <script-name> — see package.json).
• types/ — shared TypeScript types, including the canonical Powertrain schema that the database, prompts, and validation all derive from.

For project conventions, the data model, and other durable notes, see CLAUDE.md.

Deployment

Deployed on Vercel. Pushing to main triggers a production deploy automatically.