feat: add multimodal analysis support to LLM moderation client by processing image attachments

tests/media/mediaController.test.ts

@@ -225,7 +225,10 @@ describe("MediaController", () => {
     expect(screenController.start).toHaveBeenCalledWith(
       "https://youtu.be/video",
     );
-    expect(resolveMediaSource).toHaveBeenCalledWith("https://youtu.be/video", "screen");
+    expect(resolveMediaSource).toHaveBeenCalledWith(
+      "https://youtu.be/video",
+      "screen",
+    );
     expect(state).toMatchObject({ playing: true, activeMode: "screen" });
   });
 

tests/media/musicPlayer.test.ts

@@ -71,10 +71,14 @@ describe("createMusicPlayer", () => {
       ],
       { stdio: ["ignore", "pipe", "pipe"] },
     );
-    expect(discordPlayer.playStream).toHaveBeenCalledWith(proc.stdout, "music", {
-      inputType: StreamType.Raw,
-      inlineVolume: true,
-    });
+    expect(discordPlayer.playStream).toHaveBeenCalledWith(
+      proc.stdout,
+      "music",
+      {
+        inputType: StreamType.Raw,
+        inlineVolume: true,
+      },
+    );
   });
 
   it("rejects playback when Discord is not connected", () => {