Image

Generate an Open Graph / social share image (og:image, Twitter card) for the current project — a short title pitch over the product's real interface, built in the project's own stack from its real layout, components, tokens and copy, captured at 2x as a PNG and wired into the page meta tags. Use when asked for an OG image, social preview, share card, link preview image, or twitter:image.

Convert real-world photographs into Soviet constructivist military propaganda posters (the "military constructivist" look - warm cream paper, vermilion red, deep black, ochre gold, hard-edged flat color blocks, bold black ink outlines, diagonal wedges, stepped terraces, industrial skyline silhouettes, parallel hatch lines, halftone offset-print texture), and render new images in that style from scratch. Backgrounds must never use radiating sunburst rays or any fan-shaped light motif, which reads as a Japanese rising-sun flag. This skill should be used when the user supplies a photo and asks to turn it into a poster, mentions 军构美学 / 构成主义海报 / 军事宣传画 / 苏联海报风格 / 复古军事版画 / constructivist military poster / soviet propaganda poster, or wants a batch of images rendered in this style.

Generate a logo mark for a software product in the manner of Notion, Linear, Vercel, Slack, Miro, Airtable, Intercom, PostHog, Webflow, Mintlify and Claude — one flat geometric symbol, in one of the five families those marks fall into — through an image model, several concepts at once, then pick, and cut the pick into transparent PNGs ready to drop into a site: the mark alone and the app-icon tile. Use when asked for a logo, logomark, app icon, brand mark, favicon or icon for a product, \"like Linear's\" or \"like Slack's\", or a logo in the style of a named SaaS brand.

生成「产品场景图」——扁平矢量 UI 示意图风格的 1760×1120 PNG。特征为纯白画布、三层嵌套圆角卡片、蓝紫 135° 渐变、严格 0° 透视、无人物无照片。提供 v2（卡片 + 短标签）与 v3（大图标 / 圆形节点 / 图形装置，文字极少）两种取向，内置 50 枚线性图标与 15 件图形零件。核心约束是「缩略可读」：图会被放进网页卡片缩到 240px 宽显示。当需要为产品功能、方案卖点、技术能力配一张风格统一、可直接放进 PPT / 官网 / 产品页的示意图，或需要按标题批量产出同风格图示时使用。触发关键词：场景图、产品场景图、功能场景图、示意图、配图、图标化示意图、图形化配图、扁平矢量插画、flat vector product illustration、product scenario illustration。

根据用户描述的故事内容，润色故事线并拆分为分镜脚本，批量生成风格一致的故事板插图。提供8种视觉风格选择，基于故事类型推荐最佳风格。调用Gemini API生成图片，保存到Obsidian图片目录。触发词："分镜故事"、"故事板"、"做个分镜"、"创建故事板"、"故事板"、"画个故事"。

Create clean, modern logo concepts in the style of rounded app icons: soft square tiles, simple high-contrast symbols, mascot silhouettes, abstract monograms, subtle depth, and minimal typography. Use when the user asks for app icons, clean logos, mascot marks, SaaS/product logos, icon grids, logo prompts, AI-image logo generation, or a logo matching the bundled rounded-square reference examples.

Turn one character image into a 3D collectible figure with the img2fig CLI - look-dev render, four-view turnaround, view cropping, targeted image edits, multiview image-to-3D on fal (Tripo by default, optional bake-off), Blender prep (orientation, normalise, decimate, soft-vinyl grade, blush), review renders, comparison sheet, turntable video, and 3D-print repair (watertight solid, fragments, thin-wall check, hollowing, base, STL/3MF, validation report). Paid calls are dry-run first with a cost estimate and a budget guard. Use when someone wants a figure-style 3D model or turntable of a character they have the rights to use.

Turns one image of any subject (an AI image, a render, a photo of a figurine or product, a drawing, or an idea turned into an image first) into a watertight 3D sculpt that is measured against the image and recorded step by step in Blender: front maps (mask, metric depth and focal length, normals), generated views of the unseen sides (registered over the whole circle and screened), a visual hull, then many recorded fit passes against every normal map, silhouette and the photo's depth, with symmetry, focus regions, gap cuts and smoothing. Ships a CLI (`scripts/i2s.py`) for case scaffolding, fal calls with a budget, registration, carving, fitting, map checks, Blender replay, timelapse and turntable films, exports (GLB, OBJ, STL) and a report. Use when the user wants a 3D model or sculpt "from this image", a figurine or character turned into a mesh, a sculpting timelapse, or a check of a 3D model against a reference image. Not for scenes or rooms, text-to-3D without any image, rigging...

Default image workflow using a relevant Image Bible character reference and location reference. Copy their visual construction, then remix the character and exactly five location elements. Carry into photographic keyframes and other skills; explicit user requirements override the corresponding defaults.

Draw and paint figures, characters, light and colour in any style, with Krita 5.x, Blender Grease Pencil v3 or code, and check a figure or character sheet for anatomy and pose errors. Use for drawing from gesture and construction, anatomy, foreshortening, painting light and form, colouring a grayscale painting, Krita setup, hand-painting textures for 3D and turnaround or style sheets (\"anatomy\", \"Krita\").

Design, revise, or audit a logo mark on a geometric blueprint (construction grid, circles, triangles, halved angles) that anyone can rebuild without measuring, then prove it with measured tests and a variation tournament. Use when asked to create a logo, revise a logomark, explore logo variations, build a logo construction or grid diagram, make a mark crop-safe for avatars or app icons, produce positive and reversed colourways, score logo options for scalability down to favicon size, export a single-path 24px icon, or check whether a logo can be redrawn from scratch. Triggers - "logo blueprint", "logo grid", "construction lines", "geometric mark", "rebuild this logo", "logo variations", "avatar crop", "favicon test", "reverse the logo colours", "simple icon".

Make, edit or fix an animated pixel-art sprite as one SVG file: a mascot, character, animal, creature or object, with animation states the user names (default: one idle loop). Builds it from a Python builder script, renders a PNG preview, and opens a local studio that shows every state animating side by side with a pixel editor. Use when someone asks for a pixel mascot, sprite or avatar, wants an existing one to look or move differently, wants to hand-edit its pixels or watch it animate in a browser, or reports one that is frozen, misdrawn or leaves trails.

Create cloud-architecture-style diagrams of tech stacks, system architectures, infra layouts, and service topologies — with real brand logos, grouped containers, and clean orthogonal arrows — delivered as self-contained HTML, a PNG image, a React/TSX component, or a live artifact. Use this skill whenever the user asks for an architecture / system / stack / infra diagram, a tech-stack overview, a service topology, a cloud-vendor-style diagram, or wants tech logos arranged into a picture of how their services connect — in any language, even if they don't name an output format, and even if the tools involved aren't cloud services.

绘制和修改 Micro-UI 矢量插画。结合真实控件切片参考、分层构图、比例与光影参数、颜色角色映射及可编辑案例，用 SVG 制作封面、功能插画和组件展示图。

在 three.js 用程式做天空與窗外遠景，不需要 HDRI 或圖檔。天空的同一份參數同時用在三個地方：烘焙光照用的解析式天空、看得到的天空球（雲層、高空薄雲、地平線霧帶、太陽），以及不烘焙時用的太陽光、半球光與環境貼圖。內含四種預設（清晨、白天、陰天、黃昏）和一個量測工具，確認看到的天空和照亮場景的天空亮度一致。遠景部分包括：只建從室內窗戶看得到的東西（窗洞可見性與 LOD）、隨距離變淡的大氣霧化、遠處逐漸變粗的草地網格、碎石小路、樹籬、鄰居房子（陶瓦屋頂、窗戶、煙囪、天窗）、封住地平線的遠方樹牆，並可以搭配 threejs-procedural-vegetation 種樹。當使用者說「天空」「sky shader」「雲」「程式生成天空」「不用 HDRI」「窗外景色」「遠景」「背景房子」「地平線」「大氣透視」「霧化」「窗外的樹」「天空和光照對不上」「清晨／黃昏／陰天的天空」「只建看得到的東西」時使用。

用 three.js 從零做一個照片級的室內場景，完全用程式生成、不需要模型檔、貼圖或 HDRI 的總綱與起手式。內容包括：整體流程（尺寸與座標 → 外殼與窗 → 主角物件 → 材質 → 小物 → 烘焙光照 → 後製 → 驗證 → 電影輸出）、各步驟該用哪個 skill、跨 skill 整合的規則（決定性、細分、材質要經過 GI 補丁且保留原本的 shader 補丁、反照率、太陽直射與烘焙分工、陰影漏光）、一個指令就能建立可執行的新專案（房間、窗、木地板、牆面、桌椅、掛在椅背的毛巾、書架、杯子、水壺、現場烘焙、後製），以及驗收清單和實測數字。當使用者說「做一個寫實的房間／室內」「three.js 照片級場景」「程式生成整個場景」「不用模型做室內」「room-at-dawn 那種」「從哪裡開始做寫實 3D 場景」「開一個新的寫實場景專案」「把這些 skill 串起來」「寫實室內場景流程」時使用；做其中單一環節時，直接用對應的子 skill。

用程式「設計」出照片級的靜態布料（不做物理模擬）：沒鋪好的棉被、毯子、桌巾、垂下床緣或桌緣的布、翻起一角的被子、布上放東西，以及掛在桿子上的窗簾（吊環夾出的折、往下鬆開與散開、綁起下擺、車線與雙層縫邊、亞麻平織）。方法是平面高度場（團塊＋有方向的長折＋削峰＋逐條畫上的皺褶）→ 包到支撐物上 → 高的折往一側倒 → 翻角 → 依 3D 曲面攤平 UV → 細皺紋 normal map＋凹處變暗，含可重用程式庫、分階段範例、驗證數據，以及搬到 Godot／Unreal 的做法。當使用者說「程式生成布料」「棉被／床單／毯子／桌巾建模」「布料皺褶」「布垂下床邊」「被子翻一角」「布料不要用模擬」「cloth without simulation」「procedural cloth」「布上的花紋被拉長」「布料看起來像塑膠／像山」「窗簾建模」「程式生成窗簾」「窗簾皺褶」「curtain」「布簾」時使用。

在 three.js 用程式即時生成照片級材質貼圖（不用任何圖檔）：可無縫重複的雜訊、以公釐為單位的高度場轉 normal map、粗糙度與 AO 打包、Canvas2D 轉貼圖、平均反照率（給光照烘焙用）、接縫檢查。內含現成的材質：滾筒漆木屑壁紙加上大尺度的灰泥與漆色變化、烤漆橘皮紋、手摸過的粗糙度、積灰、橡木地板（交錯接縫、每片紋理不同、節疤、四邊微倒角）、通用木紋、對拼橡木貼皮、平織、鬆餅格、毛巾布、羅紋、針織、毛氈、棉繩、使用痕跡（灰塵、油脂、指紋、碎屑、棉絮、頭髮、杯印、細刮痕，同時畫進顏色與粗糙度）、斑點釉、粉體塗裝刮傷、書封圖集與對應 UV 的書本幾何、書頁邊、地形圖海報、照片、假文字。當使用者說「程式生成貼圖」「procedural texture」「不用圖檔的材質」「normal map 怎麼產生」「木地板／木紋貼圖」「牆面看起來太假」「布料材質」「針織／毛巾布 normal map」「使用痕跡」「刮痕／指紋／灰塵」「粗糙度貼圖」「書封／書背」「假文字」「貼圖有接縫」「Canvas 畫貼圖」時使用。

Açıkça çağrıldığında logo adayları geliştirir; kullanıcı seçiminden sonra gerçek fontlarla temiz vektör, düzenlenebilir Canva kaynakları ve kontrollü varyantlar hazırlar.

Orchestrate a fail-gated, source-locked Blender reconstruction of mascots, logos, brand avatars, and stylized flat characters from wireframes, texture packs, and orthographic views. Use when the user requires a 1:1 brand/model match rather than a plausible stylized interpretation.

从参考立绘通过 Tripo 兼容代理生成带贴图的 3D 人物，接续 Blender 外形检查、组件修整、按需绑骨动作与资产交付。用于人物建模和已有生成任务续作。

当用户希望将PPT或PPTX课件整理为深入易懂的图文学习笔记，包含教学配图、分章制作、审查或多种格式交付时使用。面向Codex及具备文件和图像工具的OpenAI智能体环境。

Render a Roblox item (model, tool, accessory, part) as a clean transparent-background image using only Roblox Studio viewport screenshots, by staging a copy in empty sky and shooting it against white and black backdrops. Use whenever the user wants an icon, thumbnail, shop image, inventory picture, cutout, sprite or "render" of something in a Roblox place, wants many items rendered in a batch, or mentions the Photobooth plugin's technique, even if they only say "take a picture of this sword". Requires the Roblox Studio MCP tools (execute_luau, screen_capture).

专业服装美术/时尚造型/真人写真提示词写作助手。用于为AI绘画与生图模型(Midjourney、SD、Flux、即梦、可灵、Nano Banana、GPT-image等)撰写高质量的成年人物服装、配饰、首饰、面料、廓形、姿势、场景、光线、镜头提示词,也可用于小说/漫画/游戏人设的服装描写。学习巴黎世家、华伦天奴、圣罗兰、Prada、Miu Miu、The Row、麦昆、Versace、Dior、Chanel、Bottega Veneta、Loewe、Mugler、Schiaparelli、Alaïa、Jacquemus等品牌语言,覆盖针织开衫、毛衣裙、旗袍/新中式、丝袜与鞋履、蕾丝、真丝睡裙、泳装、职业装、礼服露背等品类。支持:手机随手拍/CCD直闪/胶片等真实感写法、姿势与人体结构锁定、批量随机池、角色一致性锁定、COS写真。只要用户提到写真、人像、男装、男模、西装、男性造型、服装描写、穿搭、造型、人设、lookbook、生图提示词、"帮我写一套衣服/一张图"、"有魅力/性感/高级感/氛围感的穿着",即使没有明确说"skill"或"提示词",也要使用本skill。