Image

Use when working with Grok MCP tools (agent prompts and tool loops, image analysis, image generation), OR when estimating token cost of a task running on your current model, OR when a task involves >50K tokens, large files, or full codebase analysis — to decide whether to route through Grok for cost savings and select the right model.

Use when rendering Mermaid to SVG. Renders Mermaid 12 source to an SVG or HTML file with Pele, without a browser. Covers 32 types: flowchart, sequence, class, state, ER, gantt, git graph, mindmap, pie, journey, timeline, quadrant, xy chart, C4, block, architecture, requirement, sankey, packet, radar, treemap, kanban, railroad grammar, Wardley map, Ishikawa/fishbone, swimlane, Venn, use case, file tree, Cynefin, event modeling, agent flow. Use for .mmd files or any request for an image of one of these. Do not use to leave a Markdown fence as-is, or for a diagram that is not Mermaid.

Guide to Sonner, the React toast library — install and wire up the Toaster, pick the right toast() call, promise and loading toasts, updating, dismissing and persisting toasts, styling, theming and icons, positioning and multiple toasters. Use when working with Sonner or troubleshooting it — toasts that don't appear, appear twice, lose their styles, ignore Tailwind classes, sit behind a modal, or don't follow dark mode.

>- Produce and visually inspect Dramagic character-pack artwork: avatar, full-body signature image and four consistent transparent half-body emotion sprites. Use for generating, finding authorized sources, editing or repairing NPC pack imagery. Provider and workflow agnostic; evaluate actual files, not prompt quality.

Use when someone has an SVG (an icon, a logo, an illustration) and wants it as a Hairline figure, or runs /hairline-adopt with a file. Reads the drawing, rebuilds its object as an isometric line figure on the engine of @lucasmarkes/hairline, and gives it a gesture that answers the pointer.

Use whenever a turn suggests, gives an opinion on, designs, or implements anything that changes what users see - new or changed screens, tabs, buttons, fields, layout, copy, or user-facing features - even when the user does not name this skill, even mid-conversation in a follow-up turn, and even when the change emerges from a support/CS, research, review, debugging, or audit conversation. This includes bug fixes whose fix changes what users see (e.g. an error sheet replaced by another) and lists or tables of proposed screen changes. Also use whenever before/after images, UI screenshots for approval, or mockups are requested. Approval gate for UI/UX and feature changes - inspect the real current screen and code, get explicit approval on a side-by-side before/after preview before coding, then verify the result against it. A text-only proposal or case table is not enough. Not for merely attaching screenshots to a finished PR, and not for backend-only or invisible changes.

Review an external PDF, image, screenshot, article, or URL against a repository's current issues and code to sort its ideas into new, already tracked, shipped, contradicted, or unverifiable, read-only by default. Use when the user asks to '/check-worth' or '$check-worth' a source, or asks whether an external idea is worth pulling into a repository as an issue or code change.

Explain the big picture through actors, boundaries, and communication, building from the simplest setup to the current system. Use only when the user explicitly invokes /mental, $mental, or asks to use the mental skill. Always include a Mermaid diagram or rendered D2 image.

Create, modify, review and debug drivers for the NEEO remote through the meta engine (jac459/meta) - JSON device files with buttons, labels, sliders, switches, images, directories, listeners, discovery and registration - plus their companion services (Node-RED flows over MQTT, small local REST bridges). Use whenever the user mentions NEEO, the NEEO Brain, meta / metaCore / metadriver, a NEEO driver or device file, or wants to control a device (TV, AV receiver, media player, lights...) from a NEEO remote, or shares meta, Mosquitto or Brain logs or a Brain JSON dump to diagnose.

Create device icons for the Fibaro Home Center 3 / Yubii Home — realistic illustration of the real device on a transparent background with a small round state badge, one PNG 128×128 (+ SVG) per state, readable in light and dark mode — and upload / apply them on the HC3 (POST /api/icons, icon0…icon100, deviceIcon). Use this skill whenever the user wants icons for a QuickApp, a Z-Wave module, child devices, shutters (11 positions), sensors, or asks why an icon does not change with the state, even if they only say "make it look nicer" or "an image for my device".

Turn an HTML site (written by hand, by you, or exported from a builder) into a WordPress theme with an editable admin, using ThemeForge attributes. Use when the person wants their static or AI-generated site on WordPress, asks for a "WordPress theme from this HTML", wants the client to edit texts, images, lists, menus or products in the WordPress admin, or mentions ThemeForge. Covers the site map, placing fld-/opt-/tax-/wp- attributes, checking the markup with the free validator, and handing the site over for the build.

将 Nintendo Switch（NS）游戏适配到 Apple Silicon macOS、iPadOS 和 iOS，以 Rust 原生兼容层与 Metal 为核心，涵盖版本识别、ARM64 离线改写、Horizon 服务、NVN 到 Metal、输入音频、签名部署及实机验收。用于原生移植、跨游戏或版本适配、黑屏花屏卡死排查、移植进度评估；普通模拟器安装配置不属于本 skill。

Cria um carrossel de Instagram completo, da capa ao último slide, pronto para postar em PNG 1080x1350. Use quando a pessoa pedir carrossel, post com vários slides, conteúdo para salvar, transformar uma ideia, texto, aula ou tema em carrossel, ou melhorar a capa e a copy de um carrossel que já existe. Faz uma entrevista curta, escolhe o formato, escreve a copy slide a slide e exporta as imagens com Playwright.

>- SOLID Docker/Compose guardrails: small Dockerfiles, non-root when practical, no secrets in images, pinned base tags, clear multi-stage builds. Use when editing Dockerfile or compose files.

Local OCR using the macOS Vision framework. Triggers: OCR, 识别图片文字, 提取截图内容, 扫描件转文字, 图片转文字, 扫描版PDF转文字, read text from image/screenshot/photo/scanned PDF. Local-first and offline; prefer it over cloud OCR whenever data is sensitive or the machine is offline.

QAs a project's transactional and digest emails by rendering every template without DB or network, screenshotting at 600 and 375 px in light, dark and images-off, and checking for placeholder leaks (undefined, NaN, {{…}}), preheader, unsubscribe, alt text, absolute links, Outlook-safe buttons, viewport meta and Gmail's 102 KB clip. Use when email templates change, before sending a campaign, or when asked to check how emails look.

>- Build an Instagram carousel - the cover that earns the swipe, slide-by-slide copy, and the 1080x1350 files to upload. Use when the user says "carousel", "slides", "swipe post", "turn this into a carousel", or has a list-shaped or step-shaped idea that would die as a single image.

61 portable specialist skills with searchable discovery and an explainable deterministic task router. Free under MIT. Version 0.57.

Offline on-page SEO audit of a local HTML file — title and meta description quality, canonical, robots meta, viewport, Open Graph tags, h1 structure, image alt coverage, lang/charset. Zero network access. Use when the user asks to check a page's SEO basics from saved HTML or a template file.

Turns an approved script into scenes, blocking sheets and shots (projects/<slug>/shotlist.json) built for continuity — each scene is storyboarded as one grid image and filmed as one multi-shot generation, so characters keep their positions, props and screen sides across cuts. Use after the script is approved, when shots break continuity, or when planning coverage for dialogue.

Creates and maintains the reusable bible of characters, sets and styles (bible/), including character sheets, voice choice and voice samples, and the "clone me" / AI-influencer intake from the user's own photos and voice. Use when a story needs a character or location that doesn't exist yet, when a character drifts between shots, or when the user says "clone me", "make an AI influencer", "new character", "crea un personaje".

Give an interface one shared tempo and a small vocabulary of hover and state micro-interactions: duration tiers (tap/pop/hover/card/layout) for CSS and Motion, swap hand-offs, rows whose highlight grows in, trailing arrows, chevron nudges, card lifts without translate, scroll-edge fades, and icon/label swaps for copy and toggle buttons. Use when animations feel inconsistent, durations are scattered magic numbers, hovers only change colour, or a product feels flat or unfinished. Triggers on: animation duration, transition timing, hover state, micro-interaction, polish, copy button feedback, menu icon to close, back arrow, card hover, scroll fade, motion tokens, Framer Motion durations.

Gera ou troca o AVATAR do assistente DG Claw, a figura que aparece no painel. Use quando o usuario pedir "gerar o avatar", "criar a carinha dele", "mudar o mascote do painel", "trocar a foto do assistente", "fazer a imagem dele", ou rodar /dgclaw:avatar. Explica os tres caminhos (usar o que ja vem, gerar a partir da aparencia descrita, ou subir a propria imagem) e onde os arquivos ficam. As expressoes que mudam com o humor sao da extensao DG Alive.

Research, generate and refine logo concepts through explicit lettering structure, visual references and focused critique. Use for wordmarks, monograms, emblems, logo prompts or concept reviews; generated images remain concept studies.