Token导航 LogoToken导航

Agent Skill · 研究检索

frontend-planning(前端规划)

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。

frontend-planning 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为接触敏感数据;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 68,166。

总安装 235周安装 10GitHub Stars 2下载 82核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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-planning(前端规划)
来源仓库:https://github.com/ywj3493/claude-skills
仓库路径:skills/frontend-planning
安装命令:
npx skills add https://github.com/ywj3493/claude-skills --skill frontend-planning
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。

skills.shnpx skills
npx skills add https://github.com/ywj3493/claude-skills --skill frontend-planning

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

辅助前端功能模块拆分与技术方案设计。输出任务清单、接口契约与原型草图。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。支持敏捷迭代中的需求澄清与风险评估。建议产出物纳入项目管理工具跟踪执行。

frontend-planning 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/frontend-planning

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/ywj3493/claude-skills --skill frontend-plannin…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要接触敏感数据,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

frontend-planning 怎么安装?

在终端执行 npx skills add https://github.com/ywj3493/claude-skills --skill frontend-planning。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

frontend-planning 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

frontend-planning 需要什么权限?

权限级别是接触敏感数据。该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

frontend-planning 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

frontend-planning 用的人多吗?

公开数据是累计安装 235 次,近一周安装 10 次,下载 82 次,GitHub 2 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

frontend-planning 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 需要接触敏感数据,装到常用环境前要先评估影响面。
  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 环境里有大量密钥、凭证,不方便给第三方 Skill 读取的人。
  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
frontend-planning

前端规划 · 本页

2352接触敏感数据未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

frontend-planning

DEPRECATED: This skill has been superseded by dev-planning. Use dev-planning instead, which supports backend, frontend, and infrastructure domains in a unified pipeline with test traceability.

6-step planning document pipeline for frontend projects, organized by domain.

When to Use

  • User says "프론트엔드 기획", "frontend planning", "UI 설계", "화면 설계"
  • User says "프론트엔드 문서", "기획 문서 만들어줘", "plan frontend"
  • Starting a new frontend project that needs structured design documents
  • Adding a major frontend feature that requires multi-page planning

When NOT to Use

  • Backend-only projects (use the backend planning pipeline instead)
  • Fixing a single component or small UI tweak
  • Editing or updating existing specification documents
  • Non-UI work such as CI/CD, infrastructure, or tooling

Pipeline Overview

Step 0:   Tech stack detection
Step 0.5: Domain analysis & document discovery
Step 1:   Requirements          → <domain>/requirements/requirements.md
Step 2:   User Flows            → <domain>/requirements/user-flows.md
Step 3:   UI Spec               → <domain>/requirements/ui-spec.md
Step 4:   Use Cases             → <domain>/workflows/use-cases.md
Step 5:   Component Tree        → <domain>/workflows/component-tree.md
Step 6:   State & API Integration → <domain>/workflows/state-api-integration.md
Step 7:   README (table of contents)

Steps 1–3 are planning/design documents (what to build). Steps 4–6 are implementation documents (how to build).

For multi-domain projects, Steps 1–6 repeat per domain. Step 7 runs once.

Output Structure

docs/en/specifications/
├── architecture.md              # Project structure (Mermaid) — cross-cutting
├── infrastructure.md            # Deployment & infra (Mermaid) — cross-cutting
├── config.md                    # Environment variables
├── README.md                    # Index of all domains/documents (Step 7)
└── <domain>/
    ├── requirements/
    │   ├── requirements.md      # Step 1
    │   ├── user-flows.md        # Step 2
    │   └── ui-spec.md           # Step 3
    └── workflows/
        ├── use-cases.md         # Step 4
        ├── component-tree.md    # Step 5
        └── state-api-integration.md  # Step 6

Navigation

Every domain document includes two navigation blocks:

Top — sequential prev/next for linear reading:

> [← Requirements](requirements.md) | [UI Spec →](ui-spec.md)

First document omits "←", last document omits "→".

Bottom — full index to jump to any document:

---
> **All Documents**
> [Requirements](requirements.md) | ... | [State & API](state-api-integration.md)

The current document is shown in bold instead of a link.

Step-by-Step Instructions

Step 0: Detect Tech Stack

  1. Read package.json — framework, dependencies
  2. Read tsconfig.json — TypeScript usage
  3. Scan directory structure — src/, app/, pages/, components/
  4. Check styling and state management solutions

Summarize and confirm with the user before proceeding.

Step 0.5: Domain Analysis & Document Discovery

Domain Analysis

  1. Ask the user to describe major feature areas
  2. Propose domain groupings (e.g., auth, dashboard)
  3. Always create at least one domain directory
  4. Present proposed structure and wait for confirmation

Document Discovery

  1. Scan README.md, CLAUDE.md, all .md files under docs/en/specifications/ and docs/en/policy/
  2. Classify each by first 30 lines: requirements / user-stories / use-cases / api-spec / sequence-diagram / architecture / config / infrastructure / deployment / policy / other
  3. Present discovered documents grouped by category using @-reference format
  4. Carry confirmed document list to all subsequent steps

If no documents found, record "No project documents found" and proceed.

Step 1: Requirements

Output: <domain>/requirements/requirements.md

  1. Load template: references/requirements-template.md
  2. Load discovered docs classified as requirements, user-stories, or architecture
  3. Ask user for: project purpose, target users, core features, non-functional requirements
  4. Generate document and wait for review

Step 2: User Flows

Output: <domain>/requirements/user-flows.md

  1. Load template: references/user-flows-template.md
  2. Load Step 1 output
  3. Load discovered docs classified as use-cases or sequence-diagram
  4. For each major feature: Mermaid flowchart (happy path), alternative/exception paths
  5. Generate document and wait for review

Step 3: UI Spec

Output: <domain>/requirements/ui-spec.md

  1. Load template: references/ui-spec-template.md
  2. Load Steps 1–2 outputs
  3. Skim discovered docs classified as api-spec or architecture
  4. For each view: URL/trigger, layout structure, key components, responsive behavior, SEO (if applicable)
  5. Generate document and wait for review

Step 4: Use Cases

Output: <domain>/workflows/use-cases.md

  1. Load template: references/use-cases-template.md
  2. Load Steps 1–3 outputs
  3. Load discovered docs classified as use-cases or architecture
  4. For each use case: actor, preconditions/postconditions, main/alternative/exception flows
  5. Generate document and wait for review

Step 5: Component Tree

Output: <domain>/workflows/component-tree.md

  1. Load template: references/component-tree-template.md
  2. Load Steps 1–4 outputs
  3. Define: Mermaid graph of component hierarchy, shared vs page-specific components, TypeScript prop interfaces
  4. Generate document and wait for review

Step 6: State & API Integration

Output: <domain>/workflows/state-api-integration.md

  1. Load template: references/state-api-integration-template.md
  2. Load Steps 1–5 outputs
  3. Load discovered docs classified as api-spec — use as authoritative endpoint source
  4. Define: state strategy, store interfaces, API endpoints table, DTOs, caching, error handling
  5. Generate document and wait for review

Step 7: Generate README

Output: docs/en/specifications/README.md

After all domains complete Steps 1–6:

  • Single-domain: table of contents linking all 6 documents
  • Multi-domain: section per domain + optional per-domain README.md
  • If README exists, merge rather than overwrite
  • Include discovered documents in a Related Project Documents section

Report all generated file paths on completion.

Document Rules

  • Language: English
  • Meta block: Created, Last Modified, Status, Tech Stack, Prerequisites, Reference Documents
  • Mermaid: flowchart TD for user flows, graph TD for component trees
  • TypeScript: interface for props, stores, DTOs
  • References: Each step loads all previous outputs before generating
  • Review gate: Never proceed without user approval
  • Navigation: Every domain document has top (prev/next) and bottom (all documents) navigation
  • @-references: Use for discovered docs per @docs/en/policy/reference-convention.md; never @-prefix docs/reference/ files

平台分布

Codex

33.26%
按下载量换算27

Claude

31.11%
按下载量换算26

Cursor

17.71%
按下载量换算15

Gemini CLI

9.01%
按下载量换算7

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

接触敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills