Art & Design

Extract every piece of visitor-facing copy from a website codebase (templates, components, locale files, plus CMS/site settings like nav labels and meta description) into a client-ready landscape .docx review document with addressable Ref codes, a blank "Suggested revision" column, and a CSV manifest so approved edits can be applied back to source. Use whenever the user wants the words on a site reviewed, proofread, audited, extracted, exported or signed off — "extract the website copy for review", "get the client to review our copy", "all the text on the site in a Word doc", "copy review", "proofing document", "send the wording to the comms team", or a wording-focused content inventory. Not for structural, SEO or performance site audits.

Build a cinematic, award-level motion website or hero section (scroll-driven three.js scenes, film-led heroes, depth-photo parallax, synthesized sound, live 3D callouts) from one idea taken from the customer's own world. Use when someone asks for a premium animated landing page, a scroll story, a 3D product page, a video hero, an "Awwwards-style" or "expensive feeling" site, or wants a brief turned into a beat sheet, a grey-boxed mechanic and a dressed, accessible, fast page. Do not use for flat SaaS layouts, simple CSS fade-ins or static brochure pages.

Complete system for building high converting landing pages: intake questions, page structure, layout selection, conversion copywriting, SEO, plus strict visual rules for typography, spacing, corner radius, backgrounds, hero layout, icons, and motion. Use this skill whenever building, editing, styling, reviewing, or writing copy for ANY landing page, marketing site, web UI, page section, component, or prototype, even when the user does not mention design, conversion, tokens, or a design system. Always consult it before choosing page sections, headlines, CTAs, fonts, font sizes, spacing values, border radii, background colors, icons, or transitions.

Faithfully rebuild web UI from screenshots as one complete index.html using Tailwind CSS v4 via CDN. Covers admin dashboards, analytics screens, full pages, landing-page sections, tables, forms, settings screens, modals, drawers, and single components. Use whenever the user shares a UI screenshot or design image and asks to slice, replicate, clone, recreate, convert to HTML/Tailwind, do \"screenshot to code\", match a design pixel-perfectly, or fix differences between their current HTML and a reference screenshot. Not for designing new UI from scratch, redesigns, multi-page sites, React/Vue/framework output, or reproducing mockup and presentation backgrounds.

Build a black-and-white editorial landing page - a large serif headline, a drop-cap introduction, numbered ruled columns, feature spreads with framed figures (screenshots, an app window drawn in HTML, true numbers or terminal commands), a pull quote, questions and a closing line - as one responsive HTML file, drafted from the project's README, with a waitlist form that really signs people up or buttons to get started. Checks layout, contrast and accessibility in headless Chrome. Use when someone wants a landing page, product homepage, launch page or waitlist page that looks like a magazine rather than a SaaS template.

把 Unity（Mono）游戏或任意 .NET 程序的界面文本汉化成简体中文，并在写入前用自审计引擎检查完整性、一致性、格式与结构风险。适用于文本硬编码在 Assembly-CSharp.dll 的 #US 字符串堆、没有本地化表、且内嵌动态字体的项目。Use when localizing a Unity/.NET game to Chinese, extracting hardcoded UI strings from #US heap, patching ldstr tokens, or auditing an existing Chinese localization for missing/untranslated/inconsistent text.

Multi-model iterative review loop. Four models take turns improving and checking one answer, rotating roles (builder, skeptic, edge cases, simplifier) over rounds of 4x4, then 3x4, then 2x4 (36 turns), stopping early when the answer converges. Use when the user says "/assemble", wants several models to cross-check a solution, design, plan or piece of code, or asks for a Ralph Wiggum style loop.

Enterprise architecture knowledge base - design authority role, architecture principles, target vs transition architectures, macro integration patterns, non-functional requirements (availability, RPO/RTO, performance, security, operability, reversibility), cloud strategy and sovereignty (SecNumCloud, reversibility), technical debt, DORA ICT risk from an architecture view, and a cross-expert coherence checklist. Load when designing or reviewing a target architecture, arbitrating structural options, or writing ADRs; references cover TOGAF/ArchiMate, French urbanisation du SI and ADR formats.

DBA knowledge base for Oracle, PostgreSQL, SQL Server, Db2 LUW and MySQL/MariaDB - performance diagnosis, locking, HA/DR, backup/PITR, security/GDPR, Oracle-to-PostgreSQL migration pitfalls, release script best practices and a review checklist. Load it when diagnosing a database issue, designing HA/backup, preparing or reviewing a SQL release, or planning a migration.

Tune existing Virtuoso or Spectre circuits in a sequential closed loop: analyze each measured result and choose the next parameter set using the current ChatGPT/Codex agent, constrained Gaussian-process Bayesian optimization, or both. Use for W/L, bias, compensation, sizing, and design-variable optimization with explicit bounds, metrics, history-bound evidence, and independent validation.

Create or edit Virtuoso layouts with verified technology, PCells, pin geometry, layers, vias, matching and routing constraints; export GDS and run or interpret DRC, LVS, and PEX checks. Use for physical design and layout verification; connection issues and schematic testbench creation use the corresponding sibling skills.

Connect ChatGPT or Codex to a Cadence Virtuoso VM through virtuoso_bridge; discover Python and bridge configuration, verify connectivity, inspect installed API signatures, and diagnose SSH, CIW, or blocking-dialog failures. Use for connection and transport problems, not circuit optimization or design editing.

Explain Virtuoso workflows, menus, configurable hotkeys, ADE and ViVA usage, and Cadence documentation; help the user understand errors or choose a task-specific sibling skill. Use for human guidance and reference lookup, not automatic design mutation or parameter search.

>- Create a reusable content-illustration character and simple IP Standard Pack for a person, creator account, or brand. Use for personal IP, account mascots, brand characters, blind-box or designer-toy characters, “把我变成一个角色”, “做一个账号形象”, or when someone needs consistent character images, values, turnaround views, and reusable prompts.

Inspect, diagnose, and fix UI grid, alignment, spacing, sizing, and responsive layout issues with minimal changes (更多UI优化). Use when reviewing or modifying existing UI code or layouts while preserving design intent and functionality.

Design a new Hermes Bot from one sentence and spawn it with the create_agent tool. Role defaults, SOUL.md template, zero questions.

Build a real-time game (racing, arcade, top-down, platformer, puzzle, kids' learning, word, simple 3D) in an omg.dev Expo project made from the expo-game template. Start from the game kit; do not design an engine.

Create, review, refine, or update an Agent Skill when designing or maintaining its scope, trigger conditions, instructions, or supporting resources. Use for new skills and changes to existing skills; do not use for ordinary prompts, project documentation, or general agent configuration.

Prepare, publish, or recover canceled OxideTerm stable, beta, or GPUI preview releases by deriving changelog content from the previous tag, selecting the requested release-note detail level, running the repository version-bump script, validating channel-specific notes and fork ownership, committing, pushing, creating the correct annotated tag, and dispatching the main-scoped native package workflow. Use when the user asks to upgrade the OxideTerm version, prepare a release, write a release changelog, commit and push a release, create a release tag, republish the same version after its release workflow was canceled, or repair release assets from a completed Native Package run.

Create, revise, and QA publication-ready process flowcharts in a coherent Chinese journal-style visual system with configurable stage count, horizontal or vertical layouts, colored headers, aligned separators, filled process arrows, line icons, Microsoft YaHei typography, and PNG/SVG/PDF export. Use when a user asks to draw, redesign, standardize, or embed a process, evidence chain, decision framework, implementation workflow, or similar paper figure; the subject matter and number of stages can vary.

在 Yakui Studio 本地音乐工作台中制作、导入和人工修整音符与片段，结合真实分轨和转谱候选完成可追溯的参考改编或程序化音乐。

从学生需要获得的能力出发，反向设计课程目标、内容顺序、学习活动和评价，形成符合学情、课时及资源条件的教学安排。适用于整门课程、教学单元或单次课堂的设计与现有方案改进。

Review, convert, and improve UI code using Chakra UI v3. Use this skill whenever a user wants to review Chakra UI code for issues, convert plain HTML/CSS, Tailwind, CSS Modules, or styled-components to Chakra UI, clean up messy Chakra components, fix layout structure or token usage, or asks anything like "is this correct", "what's wrong with this", "review my component", "refactor this", "clean up", "convert", or "chakra-ify this" — even without the words "review" or "refactor". Trigger on any request to check, improve, or convert Chakra UI code, however casually phrased.

Use when the user's task involves Ant Design (antd) — writing antd components, debugging antd issues, querying antd APIs/props/tokens/demos, migrating between antd versions, or analyzing antd usage in a project. Triggers on antd-related code, imports from 'antd', or explicit antd questions.