Output format

All results are returned as a JSON array of objects. Each object represents one extracted item with named fields.

[
  { "title": "Example Post", "url": "https://...", "points": 142 },
  { "title": "Another Post", "url": "https://...", "points": 98 }
]

Viewing results

After a job completes, results are displayed in the format that best fits the data:

Table — the default for most structured data
Cards — image + title data like products, listings, or profiles
Timeline — date-based content like posts or news articles
Article — single long-form content with a title and body
Text — single text field results
Stats — numeric data displayed as metric cards
Comparison — side-by-side view when there are few records with many fields

The view is chosen automatically based on the shape of the data. Every result page also includes an activity log showing what the agent did step-by-step.

Exporting data

From the results page you can export as:

CSV or JSON — for any data
Text or Markdown — for single-record results only (articles, individual items)

Via the API

Results are available programmatically:

curl -H "x-api-key: api_xxx.yyy" \
  https://scrapespace.com/api/jobs/{jobId}

The response includes output (the result array), status, logs, and metadata.

Result quality

The agent extracts data dynamically from the page DOM — it never hardcodes or fabricates data. If a field can't be found on the page, it will be null in the output rather than made up.