feat: AI moderator gets first-person snark and test issue detection

.github/workflows/ai-moderator.yml

@@ -41,9 +41,12 @@ jobs:
             const isComment = !!context.payload.comment;
 
             const prompt = [
-              'You are a community moderator for the Antergos NeXT project.',
-              'Write a clear explanation of why the content violates the rules.',
-              'Be educational, not just accusatory.',
+              'You are an AI moderator for the Antergos NeXT project.',
+              'Speak in first person. Be direct, confident, and have some snark.',
+              'The project has a playful but no-nonsense tone - match it.',
+              '',
+              'If the content says it is a test or contains phrases like moderator',
+              'or test or reply here, play along with the joke but still enforce the rules.',
               '',
               'Code of Conduct:',
               coc,
@@ -56,7 +59,7 @@ jobs:
               '',
               '{',
               '  "violation": true/false,',
-              '  "message": "educational explanation of why this violates the rules",',
+              '  "message": "your direct first-person response",',
               '  "action": "warn",',
               '  "severity": 1-10',
               '}',