Token导航 LogoToken导航TokenDH.com
研究检索只读github未标认证来源可访问clear审计未展示

frontend-stack-selection前端堆栈选择

Agent Skill

用于辅助前端页面、组件、样式和交互逻辑的开发与维护。它适合让 Agent 生成或审查 React、Next.js、Vue、Tailwind、CSS 等相关代码,整理组件结构,或定位布局和性能问题。使用时需要结合项目现有设计系统、路由和构建方式,避免只生成孤立片段;涉及页面改动时,应配合本地预览和构建检查确认视觉效果。

总安装

318

周安装

13

GitHub Stars

公开资料未说明

下载量

103
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:frontend-stack-selection(前端堆栈选择)
来源仓库:https://github.com/5dlabs/cto
仓库路径:skills/frontend-stack-selection
安装命令:
npx skills add 5dlabs/cto --skill "frontend-stack-selection"
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

AgentSkills.tonpx skills
npx skills add 5dlabs/cto --skill "frontend-stack-selection"

简介

用于辅助前端页面、组件、样式和交互逻辑的开发与维护。

  • 适合生成或审查 React、Next.js、Vue、Tailwind、CSS 等相关代码,整理组件结构或定位布局和性能问题。
  • 使用时需结合项目现有设计系统、路由和构建方式,避免只生成孤立片段;涉及页面改动时配合本地预览和构建检查确认视觉效果。
  • 发现并安装 AI 代理的技能。
  • frontend-stack-selection 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

SKILL.md

Frontend Stack Selection

The CTO platform supports two frontend stack philosophies. Use this skill to choose the appropriate stack when generating Blaze tasks.

Quick Reference

StackBest ForRouterData LayerTablesForms
shadcn (Default)Marketing, SEO, SSR, e-commerceNext.js App RouterServer Actions + React Queryshadcn TableReact Hook Form
tanstackDashboards, real-time, local-firstTanStack RouterTanStack DB + QueryTanStack TableTanStack Form

Option A: shadcn Stack (Default)

Best for: Marketing sites, SEO-focused apps, content-heavy sites, e-commerce, SSR applications

Core Technologies:

  • Router: Next.js App Router (SSR, file-based routing, RSC)
  • Data: Server Actions + React Query for client-side caching
  • UI: shadcn/ui components (Radix primitives + Tailwind)
  • Forms: React Hook Form + Effect Schema validation
  • Animation: anime.js for motion design
  • Type System: Effect (type-safe errors, Schema, services)

Strengths:

  • Excellent SEO with server-side rendering
  • Built-in image optimization, fonts, and metadata
  • Streaming and progressive enhancement
  • Mature ecosystem with extensive documentation

Task XML Example:

<task id="1">
    <meta>
        <title>Build Marketing Landing Page</title>
        <agent>blaze</agent>
        <frontend_stack>shadcn</frontend_stack>
    </meta>
    <context>
        <overview>Create SEO-optimized landing page with lead capture form</overview>
        <stack_rationale>shadcn selected for SSR, SEO, and Server Actions</stack_rationale>
    </context>
</task>

Option B: TanStack Stack

Best for: Dashboards, admin panels, real-time apps, collaborative tools, offline-first, local-first applications

Core Technologies:

  • Router: TanStack Router (type-safe, client-first, search params)
  • Data: TanStack DB (collections, live queries, optimistic updates)
  • Query: TanStack Query (always included, server-state management)
  • Tables: TanStack Table (sorting, filtering, pagination, virtualization)
  • Forms: TanStack Form + Effect Schema validation
  • Lists: TanStack Virtual (100k+ items at 60fps)
  • Full-Stack: TanStack Start (optional, for greenfield projects)
  • UI: shadcn/ui components (compatible with both stacks)
  • Type System: Effect (type-safe errors, Schema, services)

Strengths:

  • Sub-millisecond live queries (~0.7ms for 100k items)
  • Optimistic mutations with automatic rollback
  • Type-safe routing with search param validation
  • Client-first architecture for instant UI feedback
  • Normalized collections prevent data duplication
  • Sync modes: eager, on-demand, progressive

Task XML Example:

<task id="2">
    <meta>
        <title>Build Analytics Dashboard</title>
        <agent>blaze</agent>
        <frontend_stack>tanstack</frontend_stack>
    </meta>
    <context>
        <overview>Create real-time analytics dashboard with live data updates</overview>
        <stack_rationale>TanStack selected for live queries and sub-millisecond reactivity</stack_rationale>
    </context>
</task>

Stack Selection Keywords

Use these keyword patterns to determine the appropriate stack:

PRD KeywordsRecommended StackRationale
dashboard, admin panel, data grid, data-heavytanstackTanStack Table + live queries excel at data-dense UIs
real-time, live updates, collaborativetanstackTanStack DB provides sub-ms live query reactivity
offline-first, local-first, synctanstackTanStack DB collections support multiple sync strategies
optimistic UI, instant feedbacktanstackBuilt-in optimistic mutations with rollback
marketing, landing page, brochureshadcnSSR and SEO optimization
SEO, search engine, metadatashadcnNext.js RSC with streaming
blog, content, CMSshadcnServer-side rendering for content
e-commerce, checkout, paymentshadcnServer Actions for secure transactions
forms-heavy, wizard, multi-stepEitherBoth have excellent form libraries

Frontend Stack in Task XML

When generating tasks for Blaze, include the <frontend_stack> element in the task meta:

<task id="{{task_id}}" priority="high">
    <meta>
        <title>Task Title</title>
        <agent>blaze</agent>
        <job_type>coder</job_type>
        <frontend_stack>tanstack</frontend_stack>  <!-- or "shadcn" -->
    </meta>
    <context>
        <overview>Task description</overview>
        <stack_rationale>Why this stack was selected</stack_rationale>
    </context>
    <!-- ... rest of task -->
</task>

If frontend_stack is omitted, Blaze defaults to the shadcn stack.


Mixed Stack Projects

For projects requiring both stacks (e.g., marketing site with dashboard):

  1. Separate Next.js routes: Use Next.js for marketing (/) and TanStack for dashboard (/app/*)
  2. Micro-frontends: Separate deployments communicating via shared state
  3. Progressive migration: Start with shadcn, add TanStack for specific features

Document the approach in the task context when mixed stacks are needed.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

04

需要参考平台分布和安装热度时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Claude Code

28.16%
按下载量换算29

windsurf

20.84%
按下载量换算21

trae

16.37%
按下载量换算17

OpenCode

13.49%
按下载量换算14

Codex

8.02%
按下载量换算8

Antigravity

3.54%
按下载量换算4

安全审计

暂无安全审计结果可展示。

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills