bozo-jiaodu

What This Skill Does

The bozo-jiaodu skill acts as an intelligent intermediary between a user's natural language creative description and the rigid technical requirements of AI image generation models. It specializes in camera optics and positioning, ensuring that requests for specific perspectives—such as low-angle shots, bird's-eye views, or specific side-profile angles—are accurately translated into the standardized <sks> prompting format required by advanced AI agents. By utilizing a comprehensive library of 96 predefined camera position combinations, this skill removes the guesswork from framing, enabling users to achieve professional-grade cinematography cues simply by typing their intent.

Installation

To integrate this utility into your OpenClaw environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/bozoyan/bozo-jiaodu Once installed, the skill will automatically monitor incoming prompts for keywords related to camera angles, shots, and perspectives.

Use Cases

This skill is essential for creative professionals, game asset designers, and digital artists who use AI to generate character models or environmental textures. If you are developing a character and need consistent renders from multiple angles (e.g., front-left quarter view for a 3D model base), this skill ensures your prompts follow the exact syntax necessary for consistency across the 96 standardized positions. It is particularly useful for users who know the visual result they want (e.g., 'a cinematic low-angle close-up') but do not want to memorize the specific technical syntax for their AI generator.

Example Prompts

1. "请把这个人物的图片生成一张左前侧的仰视特写镜头。"
2. "帮我转换成右侧平视的中景视角。"
3. "我想看这个建筑的背面，请输出背后的俯视远景提示词。"

Tips & Limitations

This skill operates on a strict keyword-trigger mechanism. If your input does not contain any indicators of camera position (like 'angle', 'view', 'shot', 'close-up', etc.), the skill will correctly identify that the request falls outside its scope and will alert the user to use general image editing tools instead. Always be as descriptive as possible with your direction; for example, combining both horizontal orientation (e.g., left side) and vertical elevation (e.g., high-angle) will yield the most accurate <sks> formatting from the internal 96-position library. Remember that this skill is a prompt formatter; the final visual quality is dependent on the image model itself.