my-ai-skills

Use when diagnosing central skill repository drift, especially when platform links look inconsistent, .skill-source.json metadata may be incomplete, verify.sh reports issues, or you want a one-command health check before larger maintenance.

Use when maintaining shared ui-polish references, helper scripts, demo assets, or bridge documents that support the ui-polish bundle.

Use when you need the non-negotiable ui-polish rules for UI direction, color, copy, spacing, readability, button feedback, and layout restraint before making interface decisions.

提升 Web、H5、小程序、Chrome 扩展、移动应用、Electron、Tauri 和桌面应用的界面质感、层级与一致性。用于任何涉及页面、组件、弹窗、设置页、面板、工具窗口、仪表盘或设计重构的任务，尤其当用户要求更好看、更高级、更有质感、更专业、减少 AI 味时。

Use before claiming a UI is done, polished, or good-looking to run the ui-polish blocking checks, lightweight validation, and final visual review.

UI/UX design intelligence for web and mobile. Includes 50+ styles, 161 color palettes, 57 font pairings, 161 product types, 99 UX guidelines, and 25 chart types across 10 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui, and HTML/CSS). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, and check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, and mobile app. Elements: button, modal, navbar, sidebar, card, table, form, and chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, and flat design. Topics: color systems, accessibility, animation, layout, typography, font pairing, spacing, interaction states, shadow, and gradient. Integrations: shadcn/ui MCP for component search and examples.

Use when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions

Execute autonomous multi-step research using Google Gemini Deep Research Agent. Use for: market analysis, competitive landscaping, literature reviews, technical research, due diligence. Takes 2-10 minutes but produces detailed, cited reports. Costs $2-5 per task.

Automates browser interactions for web testing, form filling, screenshots, and data extraction. Use when the user needs to navigate websites, interact with web pages, fill forms, take screenshots, test web applications, or extract information from web pages.

统一多平台全局规则管理（AGENTS/CLAUDE/GEMINI）。触发语义包含“新增一条全局规则”“新增全局规则”“把这句话写入全局规则”等明确写入指令；用于新增/修改规则并同步到 Claude/Codex/Gemini/Antigravity。若是讨论性提问（如“要不要写入全局规则？”）则不触发。

Guide for creating effective skills. Supports global creation in the source repository (~/Workspace/skills-central/packages/custom) and project-level creation (in ./.agents/skills). Use when creating new skills to extend AI capabilities with specialized knowledge, workflows, or tool integrations. 用于创建新技能、构建skill、制作自定义技能等场景。

Install or update skills from GitHub repositories. Supports global installation into the source repository (~/Workspace/skills-central) with runtime export to ~/.agents/skills, and project-level installation to ./.agents/skills. Enforces security checks before and after install/update with skill-security-guard. Use when installing from vercel-labs/agent-skills, anthropics/skills, or custom repos.

对本地或 GitHub 上的 AI skill 执行安装前安全审计，检测提示词劫持、下载执行、凭证窃取、数据外传、持久化与提权风险，并输出可拦截的风险结论（SAFE/CAUTION/REVIEW/BLOCK）。当用户提到“skill 安全检测”“安装前审计”“scan skill security”“恶意 skill”“prompt injection 检查”时使用。

Use when removing installed skills from the central repository, especially when cleaning unused community skills, deleting a whole bundle group, or ensuring platform links and installed-skills indexes are cleaned up consistently.

Use when updating installed skills from their original repositories, especially for centrally managed skills with .skill-source.json metadata. Supports single-skill updates and bundle-aware updates grouped by update_group.

将文本改写为"更像人写的论文/技术文档"风格。通过词汇替换、动词扩展、句式调整，让简洁表达变得稍显啰嗦、更解释性。 触发场景：论文润色、技术文档改写、让文本更像学术写作、降低 AI 检测率（另一种风格）、论文风格改写。 核心手法：动词短语扩展、系统性词汇替换、括号内容整合、句式微调。

通用代码质量检查。在提交代码前自动执行，根据项目类型应用对应的检查规则

执行 git commit 创建符合约定式提交规范的提交。仅当用户明确要求"提交代码"、"创建提交"、"帮我commit"、"git commit"等执行提交操作时使用。不要在仅仅讨论提交、询问提交规范、或解释如何提交时触发

一次性对中文论文、课程作业、技术文档和通用中文进行去 AI 痕迹改写，并支持读取本地 `.txt`/`.md` 文件输出新文件。 重组优化自 aigc-v1、humanizer-zh 和 baibaiAIGC，保留核心经验，但允许方法去重、顺序重排和更强的最终收口。

一次性对中文论文、课程作业、技术文档和通用中文进行去 AI 痕迹改写，并支持读取本地 `.txt`/`.md` 文件输出新文件。 保守整合自 aigc-v1、humanizer-zh 和 baibaiAIGC，尽量保留已验证的方法本体。

去除文本中的 AI 写作痕迹，使其更自然、更像人类书写。基于维基百科"AI 写作特征"指南。 触发场景：去除 AI 痕迹、去AI味、人性化文本、改写 AI 生成的文章、降低 AI 检测率、修复 AI 写作模式、让文本更自然。 识别并修复 24 种 AI 写作模式：过度强调意义、模糊归因、宣传性语言、三段式法则、AI 高频词、否定式排比等。

Write publishable technical articles from a clear topic plus a small amount of background. Use when the user wants to turn a technical practice, architecture change, tool evaluation, workflow upgrade, or incident analysis into a WeChat-ready draft and an X thread. The workflow is collaborative: decide whether research is light or deep, draft 1-2 paragraphs at a time, and ask only one direction-changing question when needed.

Use when UI text needs to be rewritten to remove AI flavor, self-explaining copy, redundant headings, or system-perspective language.

Use when designing or reviewing desktop application windows, tool panels, or system-flavored interfaces for Electron, Tauri, or other desktop shells.

Use when designing or reviewing Chrome extension popup, options, or side-panel interfaces where space is tight and interaction clarity matters more than decoration.

Use when buttons, actions, forms, or state transitions need clearer interaction feedback and tighter action hierarchy.

Use when designing or reviewing mini-program interfaces that need platform-native density, lightweight interaction, and restrained visual structure.

Use when designing or reviewing mobile app screens or H5 flows that must prioritize touch targets, information order, and small-screen readability.

Use when UI work benefits from live visual co-editing in Pencil so the user can see, review, and modify structure or layout directly on a .pen canvas.

Use when designing or reviewing web pages, dashboards, SaaS interfaces, admin panels, or browser-based product surfaces.

Use when a task is changing how an interface looks, feels, or is interacted with and you need ui-polish to route core rules, platform guidance, visual collaboration, and final review.