There are several default tags which are automatically imported to all metas. (See also here.) It seems that some of them have been also imported with the tag-info. By looking at metas of some relatively new sites it seems that the tag-info was imported for
- required tags: discussion, support, feature-request, bug
- moderator tags: community-ads, status-completed, status-bydesign, etc.
It has already been reported that the link to revision history does not work for those tags: Revision history links for wikis/excerpts for standard tags on metas (status-*, bug, discussion, etc) result in 404 error
I will add to the bug report linked above that if a tag info is edited then revision history is created. But it behaves as if the user who edited the tag actually created it. For example:
- The tag info for discussion on Mathematics Educators Meta has been edited today, but the revision history says it has been created today. You can see in Internet Archive that it existed long before.
- I have made suggested edit to the tag info for faq on Mathematics Educators Meta. I did not edit tag-excerpt only tag-wiki. Even though it was only one edit, it required two suggested edit reviews, both for tag-excerpt (which I did not touch) and for tag-wiki. If you check the revision history, it also says that both tag-wiki and tag-excerpt have been created today.
- I have made suggested edit to the tag info for faq on MathOverflow. Situation is the same as above - I did not make any changes to the tag-excerpt, but there are two reviews, one for tag-wiki and one for tag-excerpt. If you check the history, it says that the tag-info has been created two hours ago. In this specific instance it can be even checked in Internet Archive that the tag-info existed long before my suggested edit. In addition to this, while suggested edit is pending, the tag-info appears to be empty. Again this can be seen from Internet Archive - this capture was made while the suggested edit was pending. (When logged in, I saw the text with the usual: "This edit will be visible only to you until it is peer reviewed.")
- If a suggested edit is rejected, then the text is completely removed as can be seen from this experiment: Here is a rejected edit to tag-info. And you can see in the capture from Internet Archive that after the rejection the tag-wiki indeed appeared empty.
I will also add a link to related discussion in chat.
It is very likely that this problem and 404 error for revision history have the same root cause. But I thought that it might be worth pointing out that apart from that, this also causes incorrect attribution of edits and suggested edits are shown incorrectly (in fact, non-existent edit suggestions are sometimes generated) and rejection of a suggested edit my result in removal of the text of the tag-info.