Skip to main content
This recipe exports everything a key can read into one CSV file per resource. It follows every next_cursor, paces itself with the rate-limit headers, and skips resources the key has no scope for, so you can run it with any key.

What you need

Output files

Every file except brands.csv starts with a project_id column, so you can join them.

Code

Expected output

With a new key, on a workspace with two brands:
With an older key that lacks read:content, the articles line reads articles.csv skipped (Insufficient scope) instead, and the rest of the export still runs.

Tips

  • Open in Google Sheets or Excel. Files are UTF-8 with a header row. In Excel, use Data → From Text/CSV so non-English prompts keep their characters.
  • Load into a warehouse. The columns map one-to-one to API fields, so the API reference doubles as your schema documentation.
  • Visibility history isn’t a list endpoint, so it’s not in this export. Add it per brand with GET /projects/{project_id}/visibility (up to 366 days per request), as in Visibility over time.
  • Articles are the ones stored for each brand by the articles endpoints. Documents Quotable Agent writes are kept in the Library and aren’t part of this export.
  • Article HTML is left out on purpose: list responses don’t include it. Fetch GET /projects/{project_id}/articles/{article_id} for the articles whose content you need.
  • Formula guard. Text cells that start with =, +, - or @ get a leading ' so spreadsheets don’t run them as formulas. Remove that guard if you load the files into a database instead.
Last modified on November 18, 2026