メニュー
Fetches and analyzes web pages when a specific URL is provided. Use when non-developers share reference sites, API docs, or competitor pages. Complements web-research (search) with direct URL visits via WebFetch.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Sets up and maintains SQLite with Prisma in Node.js/TypeScript projects: prisma init, singleton PrismaClient pattern, schema conventions, db push, gitignore for dev.db, bcrypt for passwords. Use when working with Prisma, schema.prisma, prisma/ folder, SQLite data layer, or db.ts in TS/JS apps. For Python stacks, use SQLAlchemy + SQLite instead — do not apply this skill.
Deploys Next.js/web MVPs to Vercel (build check, env vars, vercel CLI). Use only after P0 features are done and tests pass — never during active development. Trigger when the user wants to go live, asks to put the app online, or confirms after "ready to publish?"
Step-by-step install guidance for Node.js, Python, and Git for beginners. Use when the user cannot run npm/python, says they are new, or asks how to install tools.
Next.js App Router + TypeScript + Tailwind + shadcn/ui setup, folder layout, and patterns (component size, mobile-first, Zustand when needed). Use for web app/service projects using Next.js — not for Python-only stacks. Trigger: .tsx/.ts, next.config, new web UI.
React Native + Expo + Expo Router + NativeWind setup, folder layout, and patterns (component size, FlatList, navigation). Use for mobile app projects using React Native — not for web-only stacks. Trigger: mobile app request, .tsx with View/Text, app.json, expo config, React Native components.
Python stacks for FastAPI, Streamlit, Gradio, LangChain, scraping (BeautifulSoup/Playwright), typed code and venv/requirements. Use for AI/RAG, data apps, API servers, automation — not for Node.js web MVPs. Trigger: .py, requirements.txt.
| name | web-fetch |
| description | Fetches and analyzes web pages when a specific URL is provided. Use when non-developers share reference sites, API docs, or competitor pages. Complements web-research (search) with direct URL visits via WebFetch. |
Core principle: When a URL is given, visit it directly. Don't guess what's on the page.
| Phase | Fetch When |
|---|---|
Planning (docs/biz/) | Reference site shared, competitor analysis, API doc page |
| Development | Reading API docs, checking library pages, verifying service status |
| Communication | Non-developer asks "what does this site do?" or "make it like this" |
Use the WebFetch MCP tool to retrieve page content from a URL.
WebFetchdocs/biz/): layout patterns, feature lists, user flows observeddocs/ai-dev/): API details, tech stack observationsWhen blocked, ask: "Could you take a screenshot and share it instead?"
WebSearchWebFetch