volcengine-video-generate - SKILL.md Agent Skill

name: volcengine-video-generate

description: Using volcengine video_generate.py script to generate video, need to provide filename and prompt, optional provide first frame image (URL or local path).

license: Complete terms in LICENSE.txt

Volcengine Video Generate

Scenarios

When you need to generate a video based on a text description, use this skill. It supports controlling the starting frame of the video using a first frame image (either a URL or a local file path).

Steps

Prepare the target filename (e.g., output.mp4) and a clear, specific prompt.
(Optional) Prepare the first frame image, which can be a HTTP URL or a local file path (the script will automatically convert it to Base64).
Run the script python scripts/video_generate.py <filename> "<prompt>" [first_frame]. Before running, navigate to the corresponding directory.
The script will output the video URL to the console and automatically download the video to the specified file.

Authentication and Credentials

First, it will try to read the MODEL_VIDEO_API_KEY or ARK_API_KEY environment variables.
If not configured, it will try to use VOLCENGINE_ACCESS_KEY and VOLCENGINE_SECRET_KEY to get the Ark API Key.

Output Format

The console will output the generated video URL.
The video file will be downloaded to the specified path.

Examples

Pure Text Generation:


python scripts/video_generate.py "cat.mp4" "a lovely cat"

With First Frame Image (URL):


python scripts/video_generate.py "dog_run.mp4" "a lovely dog is running on the grass" "https://example.com/dog_start.png"

带首帧图片生成（本地文件）：


python scripts/video_generate.py "my_video.mp4" "a person is running in the street" "/path/to/local/image.jpg"