siliconflow-video-gen
by lilei0311
Coding Agents & IDEsBrief
Generate videos using SiliconFlow API with Wan2.2 model.
Description
What This Skill Does
Generates videos from text prompts or static images using the SiliconFlow API and Wan2.2 model. Supports both Text-to-Video and Image-to-Video modes at ¥2 per video. API key can be set via environment variable or OpenClaw config.
Wan2.2 (14B params) produces cinematic-quality output accessible via a simple CLI script without managing a local GPU.
When to Use It
- Animating a product photo into a short clip
- Generating a cinematic scene from a script line
- Creating social media video content from text descriptions
- Turning concept art into motion video
- Prototyping video ads without a camera
Example Workflow
User asks: Generate a video of a sunset over the ocean with waves crashing
- 1
Reads SILICONFLOW_API_KEY from environment or ~/.openclaw/openclaw.json
- 2
Calls SiliconFlow API with Wan-AI/Wan2.2-T2V-A14B model and the text prompt
- 3
Polls the API until the video generation job completes
- 4
Downloads the resulting video file
A cinematic video clip of a sunset ocean scene saved locally
Requirements
- SILICONFLOW_API_KEY environment variable
- SiliconFlow account at siliconflow.cn
Installation
clawhub install lilei0311/siliconflow-video-genRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.