# Download and recover real tag history

Prepare genuine A4 and label PDFs, CSV and backup, then save and restore isolated work.

Source: https://help.smartfact.ai/tagbench/guides/save-and-restore

Audience: Operator

## Download and recover real tag history

1. **Make an informal observation tag**

   Make an informal observation tag.

   This lesson uses a synthetic item. A reviewed tag records an observation only; it never holds stock, isolates equipment or grants approval.

2. **Keep the exact item code**

   Keep the exact item code.

   Enter 000017. Scanning only fills the code; it creates no tag until you deliberately review the note.

3. **A repeated scan creates no extra tag**

   A repeated scan creates no extra tag.

   Enter the same code again. You remain in the capture step; a repeated reading never creates a new tag.

4. **Describe what needs attention**

   Describe what needs attention.

   Continue to the observation. The identifier remains exact, including its leading zeroes.

5. **Choose an observation reason**

   Choose an observation reason.

   Review storage asks someone to look at the item; it is not a stock or safety decision.

6. **Record what you saw**

   Record what you saw.

   Keep the note factual and clear. It is a local claim, not an authenticated audit.

7. **Add a readable description**

   Add a readable description.

   The description helps someone read the tag; it never changes the exact captured identifier.

8. **Declare the observed location**

   Declare the observed location.

   This location is the writer’s claim, not live inventory or proof of where the item is now.

9. **Create one reviewed tag**

   Create one reviewed tag.

   The genuine domain code freezes revision 1. Review its exact identifier and wording before downloading.

10. **Download the genuine A4 PDF**

   Download the genuine A4 PDF.

   The real shared renderer prepares revision 1 with its exact fields and local QR. A download never proves that a tag was physically printed.

11. **Download the genuine 100 by 150 millimetre tag**

   Download the genuine 100 by 150 millimetre tag.

   This short observation fits the fixed tag. Longer text is refused rather than clipped; A4 remains available. Printing is a separate action on your device.

12. **Download every tag revision as CSV**

   Download every tag revision as CSV.

   The actual CSV retains exact identifiers and local annotations. Spreadsheet-risk text is made safe; the backup keeps exact original values.

13. **Download a portable project**

   Download a portable project.

   The genuine exporter keeps tag revisions and prepared-download records. They are local claims, not proof of printing.

14. **Save isolated practice work**

   Save isolated practice work.

   The ordinary save action writes only this frame’s private database. Normal saved projects and workbench cookies remain untouched.

15. **Start new temporary work**

   Start new temporary work.

   Confirm the genuine replacement question. The isolated practice save remains until this session ends.

16. **List isolated saved work**

   List isolated saved work.

   Only this practice frame’s saved tags are listed. Normal TagBench projects are never read.

17. **Restore the actual saved tag**

   Restore the actual saved tag.

   The genuine strict project store restores the tag and its two prepared-download records.

18. **Refuse a malformed backup**

   Refuse a malformed backup.

   Duplicate JSON keys are rejected. The accepted tag and its immutable revision remain in place.

19. **Keep the accepted observation**

   Keep the accepted observation.

   Dismiss the refusal. The original reviewed tag is unchanged.

20. **Restore a prepared synthetic project**

   Restore a prepared synthetic project.

   The real importer validates the complete tag history and opens the supplied project as temporary work. Imported claims are unverified.

[Watch or practise this chapter](https://help.smartfact.ai/tagbench/demo?guide=save-and-restore&chapter=save-and-restore)