seedance-video

What This Skill Does

The seedance-video skill provides a robust interface for interacting with ByteDance’s powerful Seedance generative AI video models via the Volcengine Ark API. It enables users to transform text prompts or static images into high-quality, motion-rich video content. Whether you need cinematic footage, character animation, or frame-to-frame transitions, this skill manages the complex orchestration of API requests, polling for task completion, and automated file retrieval. It supports multiple model versions, including the top-tier Seedance 1.5 Pro, which includes advanced features like synchronized audio generation and draft-to-final workflows, ensuring that you can iterate on your creative vision with precision and efficiency.

Installation

To install this skill, use the ClawHub CLI command within your terminal environment:

clawhub install openclaw/skills/skills/asikochen/seedance-video-creation

Once installed, you must configure your authentication by setting the ARK_API_KEY environment variable. This is critical for the skill to authorize requests to the Volcengine API. You can set this permanently by adding export ARK_API_KEY="your-api-key-here" to your .bashrc or .zshrc file. The skill includes a dedicated Python CLI tool located at ~/.claude/skills/seedance-video/seedance.py, which is the recommended method for execution as it handles image base64 encoding and retries automatically.

Use Cases

1. Cinematic Content Creation: Generate high-fidelity video clips from descriptive text prompts for use in social media marketing or storytelling.
2. Image Animation: Breathe life into static photography or concept art using first-frame or first-and-last frame conditioning to control specific motion sequences.
3. Iterative Prototyping: Utilize the 'Draft' mode in Seedance 1.5 Pro to generate fast, low-cost video previews before committing to a full-resolution final render.
4. Automated Video Workflows: Integrate the CLI tool into your automated scripts to batch-process video generations from remote URLs or local directories.

Example Prompts

1. "Generate a 10-second video of a cyberpunk city street in the rain with neon lights reflecting on the pavement using Seedance 1.5 Pro."
2. "Take the image at ~/Desktop/portrait.jpg and animate it so the subject is smiling and looking slowly to the left."
3. "Create a video showing a transition from a closed flower bud to a fully bloomed lotus using the first frame as the starting point."

Tips & Limitations

• Model Selection: Always default to Seedance 1.5 Pro for audio-visual synchronization tasks. Use the Lite models if you are budget-conscious or prioritize high-speed output for simple motions.
• Image Constraints: Ensure input images are high-resolution and clearly defined; blurry frames can lead to artifacts in the generated motion.
• Error Handling: Use the provided Python CLI to monitor task IDs if a generation takes longer than expected. Always check your remaining API credits in your Volcengine account to avoid unexpected service interruptions during long renders.