Well, there’s two aspects to this:
-
generally speaking rules reported by external analyzers cannot be acted upon directly in the PR, and that’s by design (because they’re reported by external tools, see discussion in this other topic I mentioned).
-
in this case here the issue has the
roslynflag, when it fact its rule is a native SonarQube one (rule S4529). This looks fishy to me, let me ask some colleagues what could be going on.