my-skills

Use this skill when the task is specifically about AI search, AI Overviews, LLM citations, answer-engine optimization, or improving visibility in AI-generated answers. Good fit for AEO, GEO, LLMO, AI citations, zero-click AI visibility, and optimization for ChatGPT, Perplexity, Claude, Gemini, or similar systems. Do NOT use as the default for broad SEO audits or for pure schema implementation.

在创造性产品或设计工作之前使用此技能：创建新功能、构建新组件、添加新能力，或有意改变既定产品行为。在实现前先探索用户意图、需求和设计。不要用于 bug 修复、测试失败、构建失败、回归修复、恢复已有行为，或预期行为已经明确的实现任务；这些情况优先使用 systematic-debugging、test-driven-development 和 verification-before-completion。

Unified Impeccable frontend design skill. Use when the goal is to design, build, redesign, or improve a frontend interface — web pages, landing pages, dashboards, product UI, app shells, components, forms, settings, onboarding, empty states, responsive behavior, typography, spacing, layout, color, motion, UX copy, accessibility, or UI-perceived performance. Also use for Impeccable command-style requests aimed at frontend work, such as `impeccable craft`, `shape`, `init`, `document`, `extract`, `critique`, `audit`, `polish`, `bolder`, `quieter`, `distill`, `harden`, `onboard`, `animate`, `colorize`, `typeset`, `layout`, `delight`, `overdrive`, `clarify`, `adapt`, `optimize`, `live` (`teach` is only a deprecated alias for `init`). Do NOT use for backend or non-frontend code audits, docs, or reviews; live-site design-language extraction; static poster or cover art; HTML slide decks; PowerPoint / PDF / Word / spreadsheet deliverables; or specialized React/Next performance work.

遇到 bug、测试失败、构建失败、生产事故或异常行为时使用；在提出修复方案之前先找根因，不要凭直觉打补丁。 / Use for bugs, test failures, build failures, production incidents, or unexpected behavior — investigate the root cause before proposing a fix; do not patch symptoms.

在实现新功能、有意的行为变更，或非平凡 bug 修复时使用——先写一个会失败的测试，再写实现代码。例外：一次性原型、生成代码、配置文件、单行 trivial 修改、纯文案/排版/样式调整，以及没有可测行为的改动。 / Use when implementing new features, intentional behavior changes, or non-trivial bug fixes — write a failing test before implementation. Exceptions: throwaway prototypes, generated code, config files, trivial single-line patches, copy / typography / pure style edits, and changes with no testable behavior.

当确实需要与当前工作区隔离时使用，例如并行开发、临时 hotfix、PR 检查、多方案实验或执行大型实现计划。不要作为日常小修、普通 bug 修复、只读调查或单文件改动的默认流程。

在宣称工作完成、已修复或测试通过之前使用，在提交或创建 PR 之前——必须运行验证命令并确认输出后才能声称成功；始终用证据支撑断言。 / Use before claiming work is complete, fixed, or tests pass, and before commit / PR — run verification commands and confirm output before stating success; back every claim with evidence.

管理 Work Agent 平台能力。通过 `workctl schema` 发现动态产品和命令，再用结构化输出执行操作。

安装、升级、认证、发现并使用 Work Agent CLI (`workctl`)。适用于通过 `workctl schema` 发现并调用阿里巴巴国际站商家经营工具，处理店铺经营数据、广告、发品、商品优化、AI 图片/视频、旺铺、选品、物流、交易、IM、知识问答、深度研究等场景。 也适用于配置 workctl、登录认证、排查 token/cache/schema/runtime/async task/recovery 问题，编排多个 workctl 工具，或为新业务域接入 registry/discovery/detail/tools/call 协议。执行业务命令时必须以当前 `workctl schema` 和 `workctl <command> --help` 为准，不照搬旧 MCP/CLI 命令。

当已经有明确规格、需求或多步骤实现目标时使用，用来拆成可执行计划。不要作为普通 bug 修复、小改动或日常执行的默认入口；不要假设必须使用未安装的 superpowers execution skills。在 OpenCode 中，计划应交接给当前可用的 todo、task、验证命令和最小改动流程。

Extract the full design language from any website URL as the primary deliverable. Produces 8 output files including AI-optimized markdown, visual HTML preview, Tailwind config, React theme, shadcn/ui theme, Figma variables, W3C design tokens, and CSS variables. Also runs WCAG accessibility scoring. Use when the user wants to inspect, analyze, extract, or document a live website's visual style or design system — colors, fonts, spacing, shadows, radii, component patterns, motion style, design tokens, or CSS variables. Good fit for queries like 'what colors/fonts does this site use', '看看这个网站的配色 / 字体 / 间距体系', '我想拿到这个网站的设计 tokens', or '/extract-design'. Do NOT use when the goal is to build a new page or UI that merely takes another website as visual inspiration — that is implementation work and belongs to `impeccable`.

联网 / 网页操作的统一入口；由内部决策矩阵在 WebSearch、WebFetch、curl、Jina 和浏览器 CDP 等工具中选择最合适的执行方式。适合：需要登录态的站点、动态 JS 渲染页面、抓取小红书 / 微博 / 微信公众号等反爬严格平台、需要真实浏览器交互或多步骤网络工作流、以及检索本机浏览器书签/历史。单次公开 URL 抓取或纯关键词搜索可以直接用 WebSearch / WebFetch，不必先经过此 skill。 / Unified entry point for web and network tasks; the internal decision matrix selects WebSearch, WebFetch, curl, Jina, or browser CDP. Best fit when the task needs login state, dynamic JS rendering, scraping sites that block static fetches (小红书 / 微博 / 微信公众号), real browser interaction, or multi-step web workflows. A single public URL fetch or a plain keyword search can go straight to WebSearch / WebFetch without invoking this skill.

Create beautiful static visual art in .png and .pdf outputs using design philosophy. Use when the user explicitly wants a poster, cover, visual artwork, or other static design piece. Do NOT use by default for web UI, slide decks, or PowerPoint deliverables.

HTML PPT Studio — author professional static HTML presentations in many styles, layouts, and animations, all driven by templates. Use when the user's goal clearly implies a browser-delivered, presentation-style HTML artifact: reveal-style slides, a static web presentation, a shareable online deck, presenter-mode slides, 小红书图文, a multi-slide HTML report, or even a single-page HTML site that is intentionally composed like a PPT or editorial presentation. The user does not need to say “HTML” verbatim if the requested delivery style makes it obvious. Do NOT use as the default for generic PPT or presentation requests; prefer `pptx` for editing existing decks (or the external `ppt-master` workflow when generating a new PPTX from PDF/DOCX/URL/Markdown — `ppt-master` runs manually from its own repo, it is not a skill in this system) when an editable PowerPoint file is more likely.

Use this skill when a .pptx file is the explicit input or the most likely required output, or when the user clearly wants an editable PowerPoint file. Good fit for reading, modifying, updating, combining, splitting, or creating .pptx files, including requests where the user says 'PowerPoint', '可编辑 PPT', or otherwise implies Microsoft PowerPoint as the expected deliverable. Do NOT use for HTML or browser slide decks.

Use this skill when the task is specifically about schema markup, structured data, JSON-LD, schema.org markup, or rich-result eligibility. Good fit for implementing, fixing, or optimizing markup for FAQ, product, review, breadcrumb, organization, article, and similar schema types. Do NOT use as the default for broader SEO audits or AI-search optimization.

Use this skill as the default SEO diagnosis skill when the user wants to audit, review, or troubleshoot organic search performance, rankings, indexing, crawlability, on-page SEO, or technical SEO issues. Good fit for vague SEO requests like "my SEO is bad" or "help with SEO". Do NOT use when the task is specifically about schema markup / structured data or specifically about AI-search / LLM citation optimization.

React and Next.js performance and engineering-quality guidelines from Vercel Engineering. Use when the task is specifically about React/Next performance, data fetching, server/client rendering, bundle size, hydration, or performance-oriented code review/refactoring. Do NOT use as the default for purely visual UI design or normal frontend creation; prefer impeccable unless performance or React architecture quality is part of the task.

Use this skill when a Word document (.docx) is the explicit input or the most likely required output. Good fit for creating, reading, editing, restructuring, or polishing .docx files, including tracked changes, comments, formatting, images, headings, tables of contents, and template-style document work. The user does not need to say '.docx' verbatim if the request clearly implies a Word document handoff, such as a formal report, memo, letter, or document template meant for Word editing.

数字生命卡兹克（Khazix）的公众号长文写作 skill。仅当最终产物是公众号文章、长文、稿子、续写、扩写或按卡兹克风格写作时使用。不要仅因为输入是 PDF、brief、新闻链接或语音转文字就触发；如果用户想要研究报告、Word/PDF/PPT、SEO 审计、数据表格或简单摘要，应优先使用对应 skill 或普通分析流程。

Use this skill when a PDF file is the explicit input or the most likely required output. Good fit for extracting text or tables, OCR, filling forms, merging, splitting, rotating, watermarking, encrypting, decrypting, or creating PDFs. The user does not need to say “PDF” verbatim if the task clearly implies a PDF deliverable such as a filled form, exportable handout, or scanned-document workflow. Do NOT use as the default for generic writing, reporting, presentation, or document-editing tasks unless a PDF is genuinely part of the job.

Use this skill when a spreadsheet file is the explicit input or the most likely required output. Good fit for reading, editing, fixing, cleaning, charting, converting, or generating .xlsx, .xlsm, .csv, or .tsv files. The user does not need to say '.xlsx' verbatim if the request clearly implies a workbook or spreadsheet handoff with tabs, formulas, columns, charts, or CSV/XLSX delivery. Use when the final deliverable should be a spreadsheet file, not just an analysis.

横纵分析法深度研究 skill，用于系统性研究一个产品、公司、概念、技术或人物，并产出排版精美的 PDF 研究报告。仅在用户需要深度研究、横纵分析、竞品分析、系统调研或明确要研究报告时使用。不要用于简单名词解释、普通代码分析、bug 排查、公众号文章写作、SEO 审计、文档/PPT 转换或轻量问答。

Deploy applications and websites to Vercel using the normal interactive or already-authenticated Vercel workflow. Use when the user explicitly wants to deploy a project to Vercel and there is no special requirement to work through access tokens. Do NOT use as the default if the task is specifically about token-based Vercel CLI authentication or token-managed automation.

Deploy and manage projects on Vercel using token-based authentication. Use only when the task explicitly involves access tokens, `VERCEL_TOKEN`, CI/non-interactive auth, or token-managed Vercel CLI workflows. Do NOT use as the default for ordinary Vercel deployment requests when standard authenticated deployment is enough.

Distill a crush into an AI Skill. Import chat history, photos, social media, generate Relationship Memory + Persona, with continuous evolution. | 把暗恋对象蒸馏成 AI Skill，导入聊天记录、照片、朋友圈，生成 Relationship Memory + Persona，支持持续进化。

Audits the full six-layer Claude Code config stack when Claude ignores instructions, behaves inconsistently, hooks malfunction, or MCP servers need auditing. Flags issues by severity. Not for debugging code or reviewing PRs.

罗翔视角：以中国政法大学刑法学教授、法学科普创作者的思维框架看待法律问题、社会正义与人生哲学。 素材来源：B站"罗翔说刑法"频道、著作（《圆圈正义》《刑法学讲义》等）、媒体采访等。 核心模型：5个。决策启发式：9条。 触发词：「罗翔」「罗翔老师」「用罗翔的方式」「从罗翔视角」「罗翔怎么看」 局限：这不是法律咨询服务，不能替代律师的个案分析。素材以刑法/法治/人生哲学为主，民商法/行政法数据不足。 调研时间：2026年4月。