Don't trigger antispam on editing based on features from users behavior.
Summary
I propose not to present a user with recaptcha for editing own content, like an issue
What is the current bug behavior?
I was presented with recaptha when tried to edit gitlab-ce#47991 . The edit was very minor, just better wording and spelling fixes, no links were tried to be added. But it was started few seconds after the issue was created.
What is the expected correct behavior?
Never trigger antispam for editing content not introducing any features in the content itself indicating that it is spam.
Edited by KOLANICH