stack-selector
根据要求自动选择最佳技术栈。用户永远不会看到这一点。使用场景:在想法验证后启动新项目。触发器:仅内部使用,由构建管道调用。
📖 中文导读:根据要求自动选择最佳技术栈。用户永远不会看到这一点。使用场景:在想法验证后启动新项目。触发器:仅内部使用,由构建管道调用。 以下正文为该技能的原始 SKILL.md 内容(英文),上方安装命令可直接使用。
name: stack-selector description: | Automatically choose the best tech stack based on requirements. User never sees this. Use when: starting a new project after idea validation. Triggers: internal use only, called by build pipeline.
Stack Selector
Choose tech stack automatically. User doesn't decide.
Decision Tree
What type of app?
├─ Web app with auth/db
│ ├─ Needs real-time? → nextjs-supabase (Supabase realtime)
│ ├─ Heavy backend logic? → fastapi-postgres
│ └─ Default → nextjs-supabase
│
├─ API/Backend only
│ ├─ Python/ML focus? → fastapi-postgres
│ └─ Edge/serverless? → hono-drizzle
│
├─ Landing page / Marketing
│ └─ → landing-page (static)
│
└─ Unsure → nextjs-supabase (most flexible)
Templates
| Template | Stack | Best For |
|---|---|---|
| nextjs-supabase | Next.js 15, Supabase, Tailwind | Web apps, SaaS, dashboards |
| fastapi-postgres | FastAPI, PostgreSQL, SQLAlchemy | APIs, ML backends |
| hono-drizzle | Hono, Drizzle, Cloudflare | Edge, serverless |
| landing-page | Astro, Tailwind | Marketing, portfolios |
Selection Criteria
| Requirement | Template |
|---|---|
| "auth", "login", "users" | nextjs-supabase |
| "api", "backend", "python" | fastapi-postgres |
| "fast", "edge", "serverless" | hono-drizzle |
| "landing", "marketing", "simple" | landing-page |
| "dashboard", "admin" | nextjs-supabase |
| "ml", "ai", "data" | fastapi-postgres |
Usage
Called internally by /mvp:build:
- Analyze PRD from idea-validation
- Match requirements to criteria
- Select template
- Generate project from template
User sees: "Setting up your project..." (not the decision)