ericshang98

Clone any webpage into pixel-perfect, production-ready code. Extracts complete page structure using Playwright, intelligently chunks content following three core principles (mutual exclusivity, complete coverage, size control), and generates React/Next.js/Vue components in parallel using subagents. Use when user wants to clone a website, replicate a page design, convert URL to code, rebuild a webpage, or copy a website's layout.

Pinclaw 团队协作系统（Agent 化的团队）。当用户想用一条命令掌握项目——查"我的任务/今天要做什么/有什么活"、发起或指派任务、交付任务、验收或打回、看项目进度/全局状态时使用。每个人跑它都会先与团队当前状态同频，再看到分给自己的任务并逐个交付。任务和状态全部存在飞书多维表格。触发词：team-lark、项目状态、我的任务、今天要做什么、发起任务/指派任务、交付、验收、打回、看进度、团队任务、我负责啥、谁在做什么、pinclaw team、team lark。

查 Y Combinator 公司组合 + 维护一个私人 YC 认知库的 Skill,数据直接来自 YC 官方自己的公开 Algolia 索引(就是 ycombinator.com/companies 背后那个)。当用户想知道"最近的 YC 公司"、"YC portfolio"、"某 batch 有哪些公司"(如 YC S26 / W26 / Summer 2026)、"YC 里做 XX 的创业公司"、"哪些 YC 公司在招人"、"某公司是哪个 batch / 团队多大"、"YC 各批次统计"、"探索 YC"、"YC companies"、"YC startups"、"who's hiring at YC",或者想"建/刷新我的 YC 数据库"、"打开 YC 面板"、"我收藏了哪些 YC 公司"、"上次之后 YC 新增了哪些" 等任何 YC 公司组合查询或私人 YC 库操作时使用。即使用户只说"YC S26"、"最近哪些创业公司上了 YC"、"帮我搜 YC 里做 AI agent 的",只要上下文是 Y Combinator / YC / 加速器创业公司,就应触发。Skill 直接 curl YC 官方公开 Algolia 拿数据(免费、无限、无需 API Key),整理成中文 markdown,或维护本地快照 + 本地 HTML 仪表盘。**不要 undertrigger**——用户问 YC 现状而你不查,就是拿过时训练数据当今天的事实(YC 每年新增两三百家,你的记忆一定过期)。

Use when the user invokes /cheap-fable5 with a task, wants to save tokens/quota/usage limits, or asks to route work across cheaper models — plans the task and dispatches subtasks to subagents on difficulty-matched models (haiku/sonnet/fable).

Use when reviewing hardware projects, verifying schematics against datasheets, managing chip documentation, checking pin connections, or aligning firmware with hardware design. Triggers on .epro2 .kicad_sch .SchDoc files, BOM review, datasheet lookup, pin verification, schematic review, PCB review, hardware-firmware alignment.

Use when optimizing a website for AI search engines (GEO) and traditional search engines (SEO). Starts by checking for a positioning document (geo-shang.html), then audits, researches keywords, optimizes, and records changes. Triggers on "geo", "seo", "geo-shang", "AI search optimization", "why can't AI find my site", or any request to improve search visibility. Consumes significant resources — deep research, parallel subagents, exhaustive analysis.