A question-driven public encyclopedia where content is built through user curiosity. Instead of manual editing, the site grows whenever someone asks a question.

1. Ask - User enters a topic or question
2. Research - AI researches using real web sources (Tavily API)
3. Create - Individual entity pages are created first, then the answer page with backlinks
4. Link - Pages are automatically connected in a knowledge graph

┌──────────────────┐     ┌──────────────────┐     ┌─────────────┐
│  Next.js (SSG)   │────▶│     Go API       │────▶│  PostgreSQL │
│  - Tailwind      │     │  - Research      │     │             │
│  - TanStack Q    │     │  - Tavily Search │     └─────────────┘
└──────────────────┘     │  - Web Scraper   │
                         └──────────────────┘

• Go 1.22+
• Node.js 18+
• PostgreSQL 14+

cd api

# Copy env and add your keys
cp .env.example .env

# Create database and run migrations
createdb wikid
make migrate

# Run server
make dev

cd web
npm install
npm run dev

API (api/.env):

• DATABASE_URL - PostgreSQL connection string
• OPENAI_API_KEY - OpenAI-compatible API key
• OPENAI_BASE_URL - API base URL (optional)
• OPENAI_MODEL - Model to use (default: gpt-4.1)
• TAVILY_API_KEY - Tavily search API key

Frontend (web/.env.local):

• NEXT_PUBLIC_API_URL - API URL (default: http://localhost:8080)

• STORM-style research - Creates entity pages before answering questions
• Real citations - All facts backed by web sources
• Knowledge graph - Visual network of connected articles
• Auto-expansion - Background worker expands related topics
• Text search - Find articles by keyword