image-video-gen

Image Video Tool Workflow

Description

This is an agent workflow for generating images and videos. It coordinates byted-web-search, image-generate, and video-generate tools to complete the task.

Dependent Skills

• byted-web-search
• image-generate
• video-generate

Workflow

1. Understand user intent：
   • Receive the text description input by the user.
   • If the user input is a story or plot, directly call the byted-web-search tool to obtain background information.
   • If the user input is of another type (such as a question or request), first call the byted-web-search tool (call at most 2 times) to find suitable information.

1. Generate images：

   • Based on the prepared background information, call the image-generate tool to generate storyboard images.
   • After generation, return them as a Markdown image list, for example:

   ![分镜图片1](https://example.com/image1.png)
   

2. Generate videos (optional):

   • Based on the user input, determine whether it is necessary to call the video-generate tool to generate videos.
   • When returning video URLs, use a Markdown video link list, for example:

   <video src="https://example.com/video1.mp4" width="640" controls>分镜视频1</video>
   

Notes

• This skill itself has no Python execution script (there is no script under thescripts/ directory).
• It works by coordinating other atomic skills.
• In the input and output, any link URL involving images or videos,any form of modification, truncation, concatenation, or replacement is absolutely prohibited, and the integrity and accuracy of the original content must be maintained 100%.