Video Generation

name video-generation description Turn ideas or scripts into AI video generation prompts, shot plans, and model-ready video creation packages. metadata {"nanobot":{"emoji":"🎬"}} Video Generation Use this skill when the user wants to generate video content with AI video tools or needs prompts for video generation models. When To Use Use this skill when the user asks for things like: 生成视频 文生视频 图生视频 给我视频 prompt 做一段广告视频 把脚本变成视频镜头 给可灵 / Runway / Pika / Veo 用的提示词 Core Goal Transform rough creative intent into a Kling-ready execution package. When the product can directly generate video, prefer execution over explanation. Recommended Workflow Identify the video goal: ad story clip cinematic mood piece product demo food video talking-head enhancement Clarify the visual direction: subject setting action camera movement lighting style duration aspect ratio Produce a generation package that directly helps Kling generate the clip. Default Output Package When the user explicitly asks for a plan or prompt package, prefer this structure: Video Intent Visual Style Shot Breakdown Master Prompt Negative Prompt Prompt Writing Rules Good video prompts should clearly specify: main subject action scene/environment camera language lighting mood visual style pacing aspect ratio if needed Prefer dense, visual, specific language. Avoid: abstract-only descriptions conflicting camera instructions too many unrelated subjects overly long narrative paragraphs with no visual guidance Useful Camera Language Use terms like: close-up medium shot wide shot overhead shot tracking shot dolly in handheld slow push-in orbit shot rack focus Useful Style Language Use styles like: cinematic realism food commercial glossy product ad documentary handheld dreamy soft light neon night scene high-contrast editorial warm lifestyle photography For NanoKwali For this project, use this skill to help the user: turn short-video scripts into AI-generated scene prompts generate food video prompts, ad prompts, and social-media video prompts break a concept into 3-6 shots for separate generation produce Kling-first prompts that can be used immediately Response Rules If the user wants a directly playable video: call the generate_video tool do not say you are reading skills or documents do not repeat the same setup message twice do not include Runway / Pika / Veo comparison notes unless explicitly requested keep the chat response short and execution-oriented do not add a long "video direction" summary unless the user explicitly asks for a concept or prompt package If the user asks for a prompt or generation plan: make the package directly useful for Kling avoid generic theory avoid cross-model digressions unless asked Suggested Response Shapes Single-Shot Prompt: for one cinematic clip Multi-Shot Pack: for a sequence of generated shots Ad Video Pack: for product or brand videos Food Video Pack: for barbecue, restaurant, or recipe-style visual content