创意技能

花叔Design——用HTML做高保真原型、交互Demo、幻灯片、动画、设计变体探索+设计方向顾问+专家评审。根据任务embody对应专家（UX/动画师/幻灯片设计师/原型师），避免web design tropes。触发词：做原型、交互原型、HTML演示、动画Demo、设计变体、hi-fi设计、UI mockup、prototype、做个HTML页面、做个可视化、app原型、iOS原型、导出MP4/GIF、60fps视频、设计风格、设计方向、配色方案、推荐风格、选个风格、做个好看的、评审、好不好看、review this design、带解说的动画、解说视频、长视频科普、voiceover、narration、5分钟讲清楚什么是XX。需求模糊时进设计方向顾问（三套逻辑并行出3版真实视觉，HTML原生40种风格库网页20+PPT20为弹药）；另含品牌资产协议、反AI slop、Junior工作流、Tweaks变体、动画→MP4/G

从法语文本中去除AI写作痕迹，注入语气、个性和灵魂。适用于编辑、审校、重写或清理读起来像ChatGPT/Claude输出的法语内容。人性化、自然化、去AI化。检测并修复27种模式：AI词汇滥用（crucial, essentiel, notamment, par ailleurs, dans le paysage）、英语优先模型带来的英语化表达（faire du sens, adresser un problème）、系动词回避、公式化开头（À l'ère de, Dans le paysage actuel）……

去除文本中的 AI 生成痕迹。适用于编辑或审阅文本，使其听起来更自然、更像人类书写。基于维基百科的"AI 写作特征"综合指南。检测并修复以下模式：夸大的象征意义、宣传性语言、以 -ing 结尾的肤浅分析、模糊的归因、破折号过度使用、三段式法则、AI 词汇、否定式排比、过多的连接性短语。

对于任何制作、创建、编辑、动画或渲染视频、动画或动态图形的请求——包括宣传片、解说视频、带字幕的片段、标题卡、叠加层或任何合成内容——请先阅读此说明。HyperFrames 通过 HTML 渲染视频；这是入口技能，也是智能体创作或编辑视频的默认方式。它会将请求路由到正确的专业工作流程，并指向 HyperFrames 领域技能，因此在尝试任何其他视频或动画技能之前，请先阅读此说明，而不是猜测工作流程。

HyperFrames的所有动画知识——原子运动规则、多阶段场景蓝图、场景过渡、更广泛的动态设计技术，以及七种运行时适配器（默认GSAP，外加Lottie、Three.js、Anime.js、CSS关键帧、Web Animations API、TypeGPU）。适用于任何运动或动画任务：选取2-4条规则进行组合，或加载蓝图，或查询特定运行时的API（例如GSAP缓动函数/Lottie播放器/Three.js混合器）。HyperFrames原生特性：单一暂停时间线，安全跳帧，...

HyperFrames HTML 组合合约。用于组合结构、数据属性、剪辑、轨道、子组合、变量、媒体播放、确定性渲染规则以及最小可渲染项目的验证。

非动画创意指导，用于HyperFrames视频。适用于设计规范（frame.md / design.md）处理、调色板、排版、叙述、节拍规划

Use when a HyperFrames composition needs seek-safe 2D/3D keyframes, GSAP timelines, CSS keyframes, Anime.js, WAAPI, FLIP, paths, masks, SVG morph/draw, text trails, 3D depth, or `hyperframes keyframes` diagnostics. Don't use for broad scene strategy, brand design, media sourcing, captions, or general video planning.

对于任何制作、创建、生成、编辑、动画化或渲染视频、动画、动态图形、解说视频、标题卡、叠加层、带字幕视频、产品宣传片、网站视频、公关或更新日志视频、数据蒙太奇、动态海报或HyperFrames HTML合成的请求，请从此处开始。当用户希望HyperFrames创作或渲染完成的MP4/网络视频、选择工作流程，或在产品发布视频、无脸解说视频、网站转视频等之间路由时，请在其他视频或动画技能之前使用此功能。

通过结构化的发散与收敛思维，将原始想法提炼为清晰、可执行的概念。适用于想法仍模糊时、在制定计划前需要压力测试假设时，或希望在聚焦于一个方案前拓展更多选项时。触发词包括“构思”、“完善这个想法”或“压力测试我的计划”。

当用户想要创建、生成、编辑或优化用于营销的图像——博客主图、社交媒体图形、产品模型、个人资料横幅、列表视觉或品牌资产时。也适用于用户提及“AI图像生成”、“生成图像”、“创建图形”、“产品模型”、“主图”、“社交媒体图形”、“横幅图像”、“封面照片”、“个人资料横幅”、“列表截图”、“Flux”、“Flux Kontext”、“Midjourney”、“DALL-E”、“GPT Image”、“ChatGPT Images”、“Ideogram”等场景。

在RunComfy上编辑图像——该技能是一个智能路由，能将用户意图匹配到RunComfy目录中的正确编辑模型。可选择Nano Banana Edit（最多批量处理20张，默认保留身份特征）、OpenAI GPT Image 2 Edit（多语言图像内文本重写、多参考合成、布局精准）、Flux Kontext Pro（单参考高保真局部编辑）或Z-Image Turbo Inpaint（遮罩驱动的精准区域编辑）。该技能整合了每个模型的文档化提示模式，从而...

在RunComfy上编辑图像——该技能是一个智能路由，能将用户意图匹配到RunComfy目录中的正确编辑模型。可选择Nano Banana Edit（最多批量处理20张，默认保留身份特征）、OpenAI GPT Image 2 Edit（多语言图像内文本重写、多参考合成、布局精准）、Flux Kontext Pro（单参考高保真局部编辑）或Z-Image Turbo Inpaint（遮罩驱动的精准区域编辑）。该技能整合了每个模型的文档化提示模式，从而...

在RunComfy上编辑图像——该技能是一个智能路由，能将用户意图匹配到RunComfy目录中的正确编辑模型。可选择Nano Banana Edit（最多批量处理20张，默认保留身份特征）、OpenAI GPT Image 2 Edit（多语言图像内文本重写、多参考合成、布局精准）、Flux Kontext Pro（单参考高保真局部编辑）或Z-Image Turbo Inpaint（遮罩驱动的精准区域编辑）。该技能整合了每个模型文档中的提示模式，从而...

Mask-driven image inpainting on RunComfy via the `runcomfy` CLI. Routes to Tongyi MAI Z-Image Turbo Inpainting (the dedicated inpainting endpoint with mask, strength, and control-scale) and to identity-preserving edit models (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when a mask isn't available and the region must be described instead. Use for object removal, watermark removal, region replacement, blemish cleanup, and any controlled local edit where a binary mask defines the...

We need to translate the given English text into Simplified Chinese. The text describes an agent skill for image inpainting. We must preserve product names, protocol names, URLs, numbers, technical terms. The name "image-inpainting" is to be preserved if it appears in the source text? The instruction says: "Do not include the name unless it appears in the source text." The source text does not contain "image-inpainting" as a standalone term; it says "Mask-driven image inpainting" so "image inpainting" appears. But the name to preserve is "image-inpainting" with a hyphen. The source has "image inpainting" without hyphen. I think we should preserve the term as it appears in the source. The instruction says "Name to preserve: image-inpainting" but then says "Do not include the name unless it appears in the source text." So we only include it if it appears. It appears as "image inpainting" (two words). I'll translate that as "图像修复" but keep the

Mask-driven image inpainting on RunComfy via the `runcomfy` CLI. Routes to Tongyi MAI Z-Image Turbo Inpainting (the dedicated inpainting endpoint with mask, strength, and control-scale) and to identity-preserving edit models (Nano Banana 2 Edit, GPT Image 2 Edit, FLUX Kontext Pro) when a mask isn't available and the region must be described instead. Use for object removal, watermark removal, region replacement, blemish cleanup, and any controlled local edit where a binary mask defines the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

Image outpainting on RunComfy via the `runcomfy` CLI — extend a still beyond its original canvas, fill in what the camera didn't capture, change aspect ratio (square → 16:9, portrait → landscape) while preserving the original content. Routes across Nano Banana 2 Edit (default, spatial-language driven), GPT Image 2 Edit (multi-ref with reference-style matching), FLUX Kontext Pro (single-shot maximum-preservation), and the brand edit endpoints (Seedream / Dreamina / Qwen / FLUX 2). Picks the...

Codex的精英网站图像转代码技能。对于视觉重要的网页任务，它必须首先生成设计图像，深入分析这些图像，然后实现网站以尽可能匹配它们。在Codex中，它必须优先使用大尺寸、可读性强的分区特定图像，而非微小的压缩面板；为分区或细节视图生成全新的独立图像，而非裁剪旧图像；避免懒惰的生成不足，避免卡片嵌套的UI，并保持...

静态转视频指南：模型选择、运动提示与镜头移动。涵盖Wan 2.5 i2v、Seedance、Fabric、Grok Video及各自适用场景。用途：图像动画化、从静态图像创建视频、添加运动效果、产品动画。触发词：image to video、i2v、animate image、still to video、add motion to image、image animation、photo to video、animate still、wan i2v、image2video、bring image to life、animate photo、motion from image

在 RunComfy 上为任意静态图像制作动画——该技能是一个智能路由器，能将用户意图匹配到 RunComfy

We need to translate the given text from English to Simplified Chinese. The text describes a skill called "image-to-video" but the instruction says to preserve the name only if it appears in the source text. The name "image-to-video" does appear in the source? Actually the source text starts with "Animate any still image on RunComfy — this skill is a smart router..." The name "image-to-video" is not explicitly in the source text; it's given as the directory item name. The instruction says "Do not include the name unless it appears in the source text." So we should not add "image-to-video" to the translation. We only translate the text inside <text>. Also preserve product names, protocol names, URLs, numbers, technical terms. So "RunComfy", "HappyHorse 1.0 I2V", "Arena #1", "Wan 2.7", "audio_url", "Seedance 2.0 Pro" should remain as is. Translate the rest naturally. The text ends with

Animate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning...

静态转视频指南：模型选择、运动提示与镜头移动。涵盖Wan 2.5 i2v、Seedance、Fabric、Grok Video及各自适用场景。用途：图像动画化、从静态图像创建视频、添加运动效果、产品动画。触发词：image to video、i2v、animate image、still to video、add motion to image、image animation、photo to video、animate still、wan i2v、image2video、bring image to life、animate photo、motion from image

Animate any still image on RunComfy — this skill is a smart router that matches the user's intent to the right i2v model in the RunComfy catalog. Picks HappyHorse 1.0 I2V (Arena #1, native audio, identity preservation) for general animations, Wan 2.7 with `audio_url` for custom-voiceover lip-sync, or Seedance 2.0 Pro for multi-modal animation from image + reference video + reference audio. Bundles each model's documented prompting patterns so the caller gets sharper output without burning...

精英移动应用图像生成技能，用于创建优质的、原生应用屏幕概念和流程。专为iOS、Android及跨平台移动产品设计。优先考虑清晰的层级结构、舒适易读的文本、强大的多屏一致性、受控的调色板、非通用的创意方向、纹理表面、以图像为主导的构图、品味独特的自定义图标设计，以及整洁的手机模型框架。默认情况下，屏幕应展示在精致的iPhone或类似设备的微妙框架内...

精英前端图像导向技能，用于生成优质、注重转化的网站设计参考。关键输出规则——为每个部分单独生成一张横向图像。一个包含8个部分的着陆页需生成8张图像。切勿将多个部分压缩至同一张图像中。强制实现构图多样性（不总是左文右图）、背景图像自由、多样化行动号召、不同英雄区比例（巨型/中型/极简迷你）、叙事概念主线、二次阅读时刻，以及……

该技能指导创建独特、生产级的前端界面，避免通用的“AI 垃圾”美学。实现真实可运行的代码，并对美学细节和创意选择给予特别关注。