skillfed

aliyun-wan-video

This skill wraps Aliyun's Wan video generation models through the DashScope SDK, enabling both text-to-video and image-to-video workflows. It standardizes video.generate requests with support for prompt control, duration, frame rate, resolution, seed, and motion parameters across multiple Wan model variants.

Aliyun Wan Video generates videos using DashScope SDK with text or image prompts and configurable synthesis parameters.

AI-generated summary based on this skill's SKILL.md

397 34 MIT updated by cinience

Install

cinience/alicloud-skills/aliyun-wan-video · repository language: Python

git clone https://github.com/cinience/alicloud-skills
cp -r alicloud-skills/skills/ai/video/aliyun-wan-video ~/.claude/skills/aliyun-wan-video
npx skillfed install cinience/alicloud-skills/aliyun-wan-video

Frequently asked questions

AI-generated answers based on this skill's SKILL.md and metadata

How do I generate videos with Aliyun Wan models?

aliyun-wan-video wraps Aliyun's Wan video generation models through the DashScope SDK. To generate videos, you call the video.generate method with a text prompt (for text-to-video) or an image plus prompt (for image-to-video). The skill handles parameter normalization, async task polling, and response mapping automatically. You'll need a DashScope API key configured in your environment.

What video generation parameters does aliyun-wan-video support?

aliyun-wan-video standardizes video.generate requests to support prompt control, duration, frame rate (fps), resolution, seed, and motion parameters. These map to the underlying Wan model variants like wan2.6-t2v (text-to-video) and wan2.6-i2v-flash (image-to-video). The skill normalizes your input parameters across different regional model endpoints.

How does dashscope async video polling work in this skill?

aliyun-wan-video handles async video generation with automatic task tracking and polling. When you submit a video.generate request, the skill returns a task ID and polls DashScope until the video is ready. You can track generation progress without blocking, making it suitable for integration into video-agent pipelines where multiple tasks may run concurrently.

Can aliyun-wan-video generate videos from images?

Yes. aliyun-wan-video supports both text-to-video and image-to-video workflows. For image-to-video, use the wan2.6-i2v-flash model variant by providing a reference image along with your prompt. The skill maps image inputs and motion parameters to the DashScope API, enabling video generation conditioned on your reference image.

How do I integrate aliyun-wan-video into a video-agent pipeline?

aliyun-wan-video is designed for video-agent pipeline integration. It normalizes video.generate requests and handles async polling transparently, so your agent can queue multiple generation tasks and retrieve results without managing DashScope details. The skill's standardized interface lets you compose video generation into larger orchestration workflows.

What Wan model variants are available in aliyun-wan-video?

aliyun-wan-video supports multiple Wan model variants including wan2.6-t2v for text-to-video and wan2.6-i2v-flash for image-to-video. The skill handles regional model selection and endpoint configuration, allowing you to choose the appropriate variant based on your input type and performance requirements.

SKILL.md

rendered from the published skill — quoted content, verbatim

Category: provider

Model Studio Wan Video

Validation

mkdir -p output/aliyun-wan-video
python -m py_compile skills/ai/video/aliyun-wan-video/scripts/generate_video.py && echo "py_compile_ok" > output/aliyun-wan-video/validate.txt

Pass criteria: command exits 0 and output/aliyun-wan-video/validate.txt is generated.

Output And Evidence

  • Save task IDs, polling responses, and final video URLs to output/aliyun-wan-video/.
  • Keep one end-to-end run log for troubleshooting.

Provide consistent video generation behavior for the video-agent pipeline by standardizing video.generate inputs/outputs and using DashScope SDK (Python) with the exact model name.

Critical model names

Use one of these exact model strings:

(truncated - see the full file via the links below)

Read as markdown · JSON record · Browse the source repository

File tree — 6 files
skills/ai/video/aliyun-wan-video/SKILL.md
skills/ai/video/aliyun-wan-video/agents/openai.yaml
skills/ai/video/aliyun-wan-video/references/api_reference.md
skills/ai/video/aliyun-wan-video/references/sources.md
skills/ai/video/aliyun-wan-video/scripts/generate_dancing_video.py
skills/ai/video/aliyun-wan-video/scripts/generate_video.py

Related skills

Tags

text-to-video image-to-video async-polling dashscope-sdk video-synthesis model-studio prompt-engineering video-agent regional-variants motion-control