There are two different jobs here: deciding what belongs in your study edition, and putting the right answer behind each question. Finish the first before sending a whole deck to a printer.
- Bring the right records.
Use the native CSV download. Confirm which columns hold questions and answers, and whether the first record is a header.
- Make your own selection.
Compare repeated prompts across chapters. Keep a specific answer or keep meaningful variants; never let a filename decide.
- Turn the sheet before cutting.
Understand mirrored back positions, test the actual flip edge and fix text that cannot fit.
- Keep the edition, then share a copy.
Save an editable private working file. Distribute a PDF or reading copy only after checking its contents.
Start with the small example
The tool opens nine original demonstration cards from two fictional printing-workshop chapters. “Workshop start time?” occurs twice, once for the morning and once for the afternoon. That is a deliberate ambiguity, not a claim that an AI produced an error.
Open the repeated-prompt review. If the edition is for the morning group, keep the 09:00 card. If both groups need it, edit the prompts to identify the group, or explicitly keep both variants. Then flip the sheet to Answers and compare the card IDs.
Choose the right output
Edition JSON is for continuing your own work. It retains original cells, exclusions and edits. Browser print / Save as PDF is for fixed paper cards. Reading-copy HTML places each question directly before its answer in a normal reading order, with no scripts or remote assets.
A saved PDF does not preserve editing state. A saved JSON is not a safe public handout merely because filenames are omitted. The save and share notes explain that boundary.
CardQuire checks document structure and layout. You still need to check the study answers against your own sources. It does not generate, grade or certify educational content.