An advanced video generation skill utilizing the Wan 2.6 model to create cinematic content from images and maintain subject consistency using reference videos.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install wan-video
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install wan-video using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Wan 2.6 Video Generation?
Wan 2.6 is a cutting-edge video generation suite integrated into Openclaw Skills that enables developers to produce high-quality visual content. It supports both Image-to-Video (I2V) and Reference-to-Video (R2V) workflows, allowing for the animation of static images or the creation of new scenes that preserve the identity of subjects across different clips. Whether you are looking for high-fidelity 1080p cinematic output or a cost-effective Flash version for rapid prototyping, this skill provides the programmatic tools needed for modern AI video workflows.
Wan 2.6 Video Generation Use Cases
- Transforming static character portraits or landscapes into 5 to 15-second cinematic sequences.
- Maintaining subject consistency (people, animals, or objects) across multiple video shots using the R2V model.
- Creating multi-shot narrative videos with precise time-based camera and movement descriptions.
- Developing high-performance video previews for marketing or social media content using the Flash model variant.
How Wan 2.6 Video Generation Works
- Initialize a video generation task by calling the submit_task tool with the appropriate model ID and visual parameters.
- Provide the source material, such as a starting image URL for I2V or up to three reference video URLs for R2V consistency.
- The system processes the request asynchronously, returning a unique task_id to track the generation progress.
- Use the get_task tool to poll the status of the video until it reaches a completed state.
- Retrieve the final high-definition video URL from the result metadata for deployment or further editing.
Wan 2.6 Video Generation Setup
To utilize this skill within the Openclaw Skills framework, ensure your environment is configured to access the Wan 2.6 API endpoints. Basic task submission can be tested via CLI or integrated into your agent's workflow.
# Example of initiating a task via an Openclaw Skills compatible agent
# The agent will call the submit_task tool internally with the following payload:
{
"model_id": "wan/v2.6/image-to-video",
"parameters": {
"prompt": "Cinematic zoom in on a smiling subject",
"image_url": "https://example.com/source.jpg",
"resolution": "1080p"
}
}
Wan 2.6 Video Generation Data Schema & Taxonomy
The skill organizes its operations around task-based metadata and specific generation parameters:
| Feature | Parameter | Data Type | Notes |
|---|---|---|---|
| Core Input | prompt | String | Detailed movement description (max 800 chars) |
| Visual Source | image_url / video_urls | String / Array | URLs for the source image or reference clips |
| Video Specs | duration | String | Fixed intervals: 5s, 10s, or 15s |
| Output Quality | resolution | String | Choice between 720p or 1080p |
| Logical Control | multi_shots | Boolean | Enables cinematic shot-splitting for storytelling |
Wan 2.6 Video Generation Advanced Features
- Reference-to-Video (R2V) consistency: Use @Video1, @Video2, and @Video3 tags in prompts to maintain subject identity from source materials.
- Multi-shot Storytelling: Utilize time markers like [0-4s] and [4-8s] within prompts to define specific camera cuts and actions.
- Prompt Expansion: Automatic LLM-based enhancement of user prompts to achieve cinematic lighting, textures, and fluid movements.
- Background Audio Integration: Support for WAV/MP3 URLs to synchronize sound directly with the generated visual content.
- Flexible Model Selection: Choose between Standard models for maximum quality or Flash models for 50% lower resource consumption.
SKILL.md
Loading
Related Openclaw Skills
Vidu Video Generation
A high-performance AI text-to-video skill powered by Vidu Q3 Pro, capable of generating cinematic videos with synchronized audio.
Catbox Image Upload Skill
A streamlined utility that converts local image files into permanent, public URLs via Catbox.moe for seamless AI agent workflows.
Image Upload Skill
A high-performance utility to convert local images or remote URLs into public cloud-hosted links for AI agent workflows.
Nano Banana Pro Image Generation
A high-performance image generation and editing skill powered by Google's Nano Banana Pro model for photorealistic results.
Short Drama Plot Conception
An AI-powered creative engine that transforms single-sentence prompts into compelling, structured short drama plot outlines.
Short Drama Plot Conception
A creative AI agent skill designed to transform simple one-liners into 500-word dramatic plot outlines using a library of 50 satisfying narrative elements.