APEX FLOW WEB SCRAPER / REFERENCE LIBRARY
JSON, CSV or Markdown: which output should you use?
Use JSON for nested records and evidence, CSV for flat analysis in a spreadsheet, and Markdown for readable page content. Preserve the original structured result when exporting.
JSON preserves structure
A result can contain arrays, metadata, field evidence and validation errors. JSON keeps these relationships more clearly than a single spreadsheet cell. It is the best starting point for an application or assistant consumer.
CSV needs a defined shape
Decide which fields become columns and how arrays are represented. Spreadsheet applications may interpret cells beginning with formula characters. Apex escapes formula-like CSV cells in its export path, and tests that behavior.
Markdown serves readers
The engine extracts Markdown from supported HTML content. Markdown is useful for reading and retrieval, but it should not replace structured fields when exact values matter. The public demo also provides a Markdown sample download.
Do not lose context
Keep source_url, fetched_at and validation status in whatever format you choose. A clean-looking spreadsheet with missing context can be less useful than a verbose but inspectable JSON result.
← All reference articles · Current capabilities and limits →