Sora 2 视频生成

What This Skill Does

The Sora 2 skill for OpenClaw AI integrates OpenAI's powerful video generation capabilities directly into your agentic workflow. It enables users to create high-quality, motion-rich videos with synchronized audio from simple natural language descriptions or existing images. Whether you need a text-to-video production, an image-to-video transformation, or a creative video remix, this skill provides a streamlined interface to interact with various Sora 2 models, including standard 720p versions and high-definition 1080p Pro editions.

Installation

To integrate this skill into your environment, use the OpenClaw command-line interface: clawhub install openclaw/skills/skills/hexiaochun/sora-2 Ensure your OpenClaw agent has the necessary API credentials configured to authorize requests to the underlying fal-ai infrastructure.

Use Cases

• Content Creation: Generate b-roll, cinematic footage, or storyboards for video production using descriptive text prompts.
• Visual storytelling: Animate static assets like concept art, character portraits, or product photos into short, dynamic video clips.
• Video Refinement: Utilize the remix functionality to iterate on generated footage, changing the atmosphere, lighting, or artistic style of existing videos without starting from scratch.
• Advertising: Quickly produce multiple variations of high-resolution video content for marketing campaigns with varying aspect ratios.

Example Prompts

1. "帮我生成一个 8 秒的视频，内容是城市日落延时摄影，要求电影级画质，光线要从金黄转为粉紫。"
2. "将这张上传的图片转化为一段4秒的视频，让图中人物微笑并轻轻转头，保持自然。"
3. "针对刚才生成的那个视频，请使用重混功能，把背景切换成霓虹闪烁的赛博朋克风格。"

Tips & Limitations

• Token & Credit Usage: Video generation is resource-intensive. Pro 1080p models consume 200 points per second. Always monitor your usage to avoid unexpected costs.
• Prompt Engineering: Sora 2 performs best when you include stylistic keywords (e.g., "cinematic," "slow motion," "4K," "high detail"). Focus on describing the subject, movement, and camera angle clearly.
• Resolution: For social media platforms, select the correct aspect ratio (9:16 or 16:9) during task submission to avoid cropping issues. Standard tasks default to 720p, while Pro tasks default to 1080p. Keep in mind that longer durations result in higher credit costs linearly.