en un clic
dotfiles
dotfiles contient 5 skills collectées depuis yujiachen-y, avec une couverture métier par dépôt et des pages de détail sur le site.
Skills dans ce dépôt
Renders the response as a self-contained, visually polished HTML page saved to /tmp and opened in the browser, replacing plain markdown or wall-of-text output. Triggers when the user says "/view-as-html", or asks to "view as html / show as a page / make a pretty page / 用 html 给我看 / 做成网页 / 别给我干巴巴的文字 / 我想浏览不是阅读". Also applies when the response is structured information — comparisons, plans, status reports, research summaries, concept explainers, PR write-ups, design tokens, dashboards — even without an explicit "html" keyword, since a layout with cards, tables, SVG and color hierarchy reads faster than markdown for these shapes.
Audits an existing skill against the official Anthropic Agent Skills best practices document (https://platform.claude.com/docs/en/agents-and-tools/agent-skills/best-practices), then produces a categorized report — Must Fix (explicit warnings violated), Should Consider (acceptable but improvable), and Already Passing — with file:line citations and concrete fixes. Triggers when the user says "/review-skill", "review this skill", "audit my skill", "check this skill against best practices", "is this skill following best practices", "对这个 skill 做个 review", "审查一下这个 skill", "skill 是不是符合最佳实践". Also applies when the user references a skill directory (e.g. `agents/skills/foo` or `~/.claude/skills/bar`) and asks for sanity check, quality pass, or evaluation against documented best practices — even without naming the practices doc explicitly.
把 AI 风格的草稿改写成 yujiachen 的个人写作风格——克制 emoji、紧凑短段落、概率词代替绝对化、中性陈述代替情绪词、英文术语空格分隔。 当用户给出中文 / 中英混用稿件(小红书 / newsletter / 工程博客 / 长文评论)并要求"按我的风格改 / 去 AI 味 / 润色 / 改改这篇"等改写意图时触发;即使措辞模糊("帮我改改"),只要上下文是结构化稿件就应当考虑。
Initialize any repository with comprehensive lint rules and code quality tooling. Covers cyclomatic complexity, function/file length limits, parameter counts, type safety, code duplication, test coverage, and folder file-count limits. Supports TypeScript, Go, Python, Rust, and Java. Use this skill whenever the user asks to set up lint rules, initialize code quality checks, configure a linter, or bootstrap a new repo with quality gates — even if they don't say "lint" explicitly (e.g., "set up this repo", "add quality checks", "configure eslint/golangci-lint/ruff"). Also trigger when the user says /lint-init.
This skill should be used when exporting, storing, or reapplying agent-browser cookies across sessions, especially for localhost login states.