fix(canvas): 媒体生成 - 修复已有图片引用和图片音频视频模型匹配

web/test/canvas-model-policy.test.ts

@@ -216,10 +216,26 @@ describe("逻辑模型选择", () => {
 
     test("音频仅在单独参考时归类为音频生视频", () => {
         expect(inferVideoOperation({ textCount: 1, imageCount: 0, videoCount: 0, audioCount: 1, characterCount: 0 })).toBe("audio_to_video");
-        expect(inferVideoOperation({ textCount: 1, imageCount: 1, videoCount: 0, audioCount: 1, characterCount: 0 })).toBe("reference_to_video");
+        expect(inferVideoOperation({ textCount: 1, imageCount: 1, videoCount: 0, audioCount: 1, characterCount: 0 })).toBe("image_to_video");
         expect(inferVideoOperation({ textCount: 1, imageCount: 0, videoCount: 1, audioCount: 1, characterCount: 0 })).toBe("reference_to_video");
     });
 
+    test("图片加音频可匹配支持图生视频和参考音频的模型", () => {
+        const config = policyConfig();
+        const imageModel = "relay::cinema-image";
+        const imageCost = config.channels[0]?.modelCosts?.find((item) => item.model === "cinema-image");
+        if (!imageCost?.capabilityConfig?.video) throw new Error("缺少图生视频能力配置");
+        imageCost.capabilityConfig.video.references.maxAudios = 1;
+        const requirements = {
+            capability: "video" as const,
+            input: { textCount: 0, imageCount: 1, videoCount: 0, audioCount: 1, characterCount: 0 },
+            videoSeconds: "6",
+        };
+
+        expect(modelCompatibilityError(config, imageModel, requirements)).toBe("");
+        expect(resolveCompatibleModel(config, "relay::cinema-text", requirements)).toBe(imageModel);
+    });
+
     test("逻辑视频模型将 720 与 720p 视为同一分辨率", () => {
         const model = "cinema-720p";
         const channel: ModelChannel = {

web/test/canvas-node-generation-mentions.test.ts

@@ -32,6 +32,31 @@ function connection(fromNodeId: string): CanvasConnection {
 }
 
 describe("canvas node generation position mentions", () => {
+    test("已有图片节点显式引用自身时作为图生图参考图提交", () => {
+        const source = node("image-self", CanvasNodeType.Image, "data:image/png;base64,a");
+        source.metadata.composerContent = "将 @图片1 图片变清晰";
+
+        const context = buildNodeGenerationContext(source.id, [source], [], source.metadata.composerContent, []);
+
+        expect(context.referenceImages.map((image) => image.id)).toEqual([source.id]);
+        expect(context.imageCount).toBe(1);
+        expect(context.prompt).toBe("将 图片1 图片变清晰");
+    });
+
+    test("已有图片节点未显式引用自身时不自动退化为图生图", () => {
+        const source = node("image-self", CanvasNodeType.Image, "data:image/png;base64,a");
+        const context = buildNodeGenerationContext(source.id, [source], [], "生成一个新的构图", []);
+
+        expect(context.referenceImages).toEqual([]);
+        expect(context.imageCount).toBe(0);
+        expect(context.prompt).toBe("生成一个新的构图");
+    });
+
+    test("无法解析的画布引用会阻止静默降级为文生图", () => {
+        const target = targetNode();
+        expect(() => buildNodeGenerationContext(target.id, [target], [], "将 @图片1 图片变清晰", [])).toThrow("@图片1 没有对应的画布资源");
+    });
+
     test("同一个 @图片1 在换线后自动指向新的第一张图片", () => {
         const target = targetNode();
         const imageA = node("image-a", CanvasNodeType.Image, "data:image/png;base64,a");

web/test/generation-task-local.test.ts

@@ -1325,7 +1325,7 @@ test("remote image video and audio references keep Backend parity without Dreami
                 referenceAudios: [{ id: "remote-image-audio-audio-0001", name: "reference.mp3", type: "audio/mpeg", url: "", storageKey: "resource:remote-image-audio-audio-0001" }],
             },
             result: { mode: "video", video: { dataUrl: "opaque://video", storageKey: "resource:remote-image-audio-output" } },
-            operation: "reference_to_video",
+            operation: "image_to_video",
         },
         {
             mode: "audio" as const,