# Docs

How to call Arcadium Data scrapers from code or from an AI agent.

The API and the MCP server open at launch. Endpoints answer from launch day; until then these pages are for planning your integration.

```sh
curl -X POST https://scrapers-api.arcadiumdata.io/v1/scrapers/booking-reviews/runs \
  -H "Authorization: Bearer $ARCADIUM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "startUrls": [{ "url": "https://www.booking.com/hotel/gb/the-savoy.html" }],
      "maxReviewsPerHotel": 20
    },
    "wait": 60
  }'
```

One key, one request per run, rows back in the same call for small runs.

| Page | What it covers |
| --- | --- |
| [Quick start](/docs/quickstart/) | Your first run, from key to rows |
| [Authentication](/docs/authentication/) | API keys and how to send them |
| [Runs and results](/docs/runs/) | Starting runs, waiting for them, reading rows, errors |
| [MCP setup](/docs/mcp/) | Connecting Claude, Cursor, Codex and n8n |

## Machine-readable files

- [/llms.txt](/llms.txt): an index of this site for language models.
- [/catalogue.json](/catalogue.json): every scraper with its input schema, output fields, sample row and price.
- Every page has a Markdown version at the same address with `index.md` added, for example [/scrapers/reviews/index.md](/scrapers/reviews/index.md).