Getting Started

Requirements

• Node.js 22+
• A running Crawl4AI instance
• Anthropic API key
• SerpAPI key (for search)

Install

git clone <repo>
cd scrappy
npm install
cd ui && npm install && npm run build && cd ..

Configure

cp .env.example .env

Edit .env:

ANTHROPIC_API_KEY=sk-ant-...
SERPAPI_KEY=your-serpapi-key
CRAWL4AI_BASE=http://localhost:11235   # sets the default, can be changed in UI
PORT=3000

Start the server

npm run server

Open http://localhost:3000 in your browser.

Run your first index job

1. Go to the Scrape screen
2. Select or create a schema
3. Enter a research topic (e.g. Swiss Säule 3a accounts)
4. Click Start Index
5. Switch to Monitor to watch the agent work in real time

CLI alternative

npm start -- index \
  --topic "Swiss savings accounts" \
  --schema schemas/3a-konto.ts \
  --output my-dataset \
  --max-iterations 40