@mindstudio-ai/agent 0.1.152 → 0.1.153

package/dist/index.d.ts

@@ -10267,10 +10267,12 @@ interface StepMethods {
      */
     generateStaticVideoFromImage(step: GenerateStaticVideoFromImageStepInput, options?: StepExecutionOptions): Promise<StepExecutionResult<GenerateStaticVideoFromImageStepOutput>>;
     /**
-     * Generate a video from a text prompt using an AI model.
+     * Generate a video from a text prompt, an image or references, using an AI model.
      *
      * @remarks
-     * - Prompts should be descriptive but concise (roughly 3–6 sentences).
+     * - Prompts should be descriptive but concise (roughly 3–6 sentences): the subject, the motion, the camera and the sound. A prompt longer than the model reads is refused, not cut.
+     * - The default model is wan3.0-video. Options go in `videoModelOverride.config` under the model's own names, and differ by model, so look the model up before writing the call. Images, videos and audio go there too, as URLs: a first frame is `image` on most models, `image_url` on Veo, gemini-omni-flash, Runway's and minimax-h3-max, `source_image` on Grok Imagine, `start_frame` on ray-2 and `first_frame_image` on skyreels-v4.
+     * - A model that makes video more than one way takes `$mode` in its config (`text-to-video`, `image-to-video`, `reference-to-video`, `video-edit` and others it lists). Left out, a first frame or reference images in the config pick their mode, and otherwise the model's default runs.
      * - Videos are hosted on the public CDN by default, or in one of the app's own file stores when the call passes `store` — a private store keeps the video private, with no public copy left behind.
      * - In foreground mode, the video is displayed to the user. In background mode, the URL is saved to a variable.
      * - When generateVariants is true with numVariants > 1, multiple videos are generated in parallel.

package/dist/index.js

@@ -4503,8 +4503,8 @@ var stepMetadata = {
   },
   "generateVideo": {
     stepType: "generateVideo",
-    description: "Generate a video from a text prompt using an AI model.",
-    usageNotes: "- Prompts should be descriptive but concise (roughly 3\u20136 sentences).\n- Videos are hosted on the public CDN by default, or in one of the app's own file stores when the call passes `store` \u2014 a private store keeps the video private, with no public copy left behind.\n- In foreground mode, the video is displayed to the user. In background mode, the URL is saved to a variable.\n- When generateVariants is true with numVariants > 1, multiple videos are generated in parallel.\n- In direct execution, foreground mode behaves as background, and userSelect variant behavior behaves as saveAll.",
+    description: "Generate a video from a text prompt, an image or references, using an AI model.",
+    usageNotes: "- Prompts should be descriptive but concise (roughly 3\u20136 sentences): the subject, the motion, the camera and the sound. A prompt longer than the model reads is refused, not cut.\n- The default model is wan3.0-video. Options go in `videoModelOverride.config` under the model's own names, and differ by model, so look the model up before writing the call. Images, videos and audio go there too, as URLs: a first frame is `image` on most models, `image_url` on Veo, gemini-omni-flash, Runway's and minimax-h3-max, `source_image` on Grok Imagine, `start_frame` on ray-2 and `first_frame_image` on skyreels-v4.\n- A model that makes video more than one way takes `$mode` in its config (`text-to-video`, `image-to-video`, `reference-to-video`, `video-edit` and others it lists). Left out, a first frame or reference images in the config pick their mode, and otherwise the model's default runs.\n- Videos are hosted on the public CDN by default, or in one of the app's own file stores when the call passes `store` \u2014 a private store keeps the video private, with no public copy left behind.\n- In foreground mode, the video is displayed to the user. In background mode, the URL is saved to a variable.\n- When generateVariants is true with numVariants > 1, multiple videos are generated in parallel.\n- In direct execution, foreground mode behaves as background, and userSelect variant behavior behaves as saveAll.",
     inputSchema: { "type": "object", "properties": { "prompt": { "type": "string", "description": "Text prompt describing the video to generate" }, "intermediateAsset": { "type": "boolean", "description": "When true, the asset is created but hidden from the user's gallery (tagged as intermediate)" }, "videoModelOverride": { "type": "object", "properties": { "model": { "type": "string", "description": "Video generation model identifier" }, "config": { "type": "object", "properties": {}, "required": [], "description": "Additional model-specific configuration" } }, "required": ["model"], "description": "Optional model configuration override. Uses the workflow's default video model if not specified" }, "generateVariants": { "type": "boolean", "description": "Whether to generate multiple video variants in parallel" }, "numVariants": { "type": "number", "description": "Number of variants to generate (max 10)" }, "addWatermark": { "type": "boolean", "description": "Whether to add a MindStudio watermark to the generated video.\n\nThe watermark is a product mark on assets generated in the MindStudio UI. A step run directly via the SDK or API is producing the app's own asset, so it is never watermarked and this flag has no effect there." } }, "required": ["prompt"] },
     outputSchema: { "type": "object", "properties": { "videoUrl": { "anyOf": [{ "type": "string" }, { "type": "array", "items": { "type": "string" } }] } }, "required": ["videoUrl"] }
   },