Archive stale pages

Move pages nobody has touched in years out of search and page trees — but keep them restorable. archive sends pages to the Confluence native archive (not the trash).

Before you start

Your site needs Confluence Standard or above. On Free, Shell stops and suggests a same-space move instead (find … | mv ./Archive).
Shell archives one page at a time (Confluence runs one archive job per site). Expect about a few seconds per page — up to a few minutes for 100.
At most 100 pages per run.

1. Find the candidates

$ cd YourSpace
$ find . -mtime +730 -prune
$ find . -mtime +730 -prune --format csv

-mtime +730 = not updated for about two years. -prune keeps the top-most match in each subtree. Save the CSV if you want a record of what you archived.

2. Preview, then archive

$ find . -mtime +730 -prune | archive -n
$ find . -mtime +730 -prune | archive -y

Or archive named pages:

$ archive -n ./Old-runbook ./Retro-2023
$ archive -y ./Old-runbook ./Retro-2023

Press Esc or Ctrl+C to stop. A page already submitted may still finish on the server.

Combine with labels

Label first, review, then archive what is still stale:

$ find . -mtime +730 -prune | label add archive-candidate -y
$ find -label archive-candidate | archive -n

Notes

Archived pages are restored in Confluence; Shell has no unarchive.
Whether child pages are archived with their parent is decided by Confluence — check the result there.
Spaces and folders can’t be archived.
Prefer the trash? Use rm (find … | rm -n). Prefer to keep pages visible? Use mv into an Archive page in the same space — see Find pages not updated recently.