What is vibe-creating-skill?

Vibe Creating is an open-source skill that rewrites your rough ideas into polished prompts for AI video-generation models. Instead of struggling to describe camera angles, lighting, and shot lists, you just describe the scene or feeling you want, and the skill translates that into language that video AI models understand well. The philosophy is that modern video models are smart enough to figure out the technical details, you just need to tell a good story. The skill first judges whether your input fits this approach. If you've written a highly specific shot list that you genuinely need, it leaves it alone. But if your idea is more about mood and emotion, it strips away low-value technical parameters and rewrites your prompt to focus on story, atmosphere, and visual intent. Every response comes back in a consistent format: its judgment, the action it took, the rewritten prompt, and notes on what it changed. This is useful for anyone creating AI-generated video, marketers, content creators, filmmakers experimenting with AI, or hobbyists. For example, if you write "a kid hits a buzzer-beater three," the skill expands that into a vivid, model-ready prompt. Or if you describe "the feeling of freedom," it recognizes that's too abstract and asks you a quick clarifying question before proceeding. What makes the project notable is that it's a single portable file following an open standard, so it works across multiple AI coding assistants like Claude Code and Codex, or you can paste it into any chatbot as a system prompt. It supports both English and Chinese, and the README includes real before-and-after video comparisons showing how rewritten prompts produce better clips.