Token导航 LogoToken导航TokenDH.com
前端设计操作浏览器github未标认证来源可访问clear审计通过

electron-architectElectron 架构师

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

2,301

周安装

94

GitHub Stars

公开资料未说明

下载量

744
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

3

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:electron-architect(Electron 架构师)
来源仓库:https://github.com/tomlord1122/tomtom-skill
仓库路径:skills/electron-architect
安装命令:
npx skills add https://github.com/tomlord1122/tomtom-skill --skill electron-architect
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/tomlord1122/tomtom-skill --skill electron-architect

简介

electron-architect 专注 Electron 应用架构设计,包括主/渲染进程划分和 IPC 规划。

  • 适用于复杂桌面应用的结构设计,提供安全通信和资源管理方案。
  • 结合平台特性(macOS/Windows/Linux)优化用户体验和性能表现。
  • 设计时需明确功能边界,避免过度耦合和安全隐患。
  • 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。

SKILL.md

Electron Architecture Expert

Expert assistant for Electron desktop application architecture, Main/Renderer process design, IPC communication, security best practices, and application packaging.

Thinking Process

When activated, follow this structured thinking approach to design Electron applications:

Step 1: Application Requirements Analysis

Goal: Understand what the desktop application needs to accomplish.

Key Questions to Ask:

  • What is the core functionality? (editor, dashboard, utility, media)
  • What system resources are needed? (file system, network, hardware)
  • What is the target platform? (macOS, Windows, Linux, or all)
  • Are there offline requirements?
  • What is the expected data sensitivity? (local files, credentials, user data)

Actions:

  1. List all features requiring system access (files, network, native APIs)
  2. Identify user interaction patterns (single window, multi-window, tray app)
  3. Determine data persistence needs (local storage, SQLite, file system)
  4. Map integration points (external APIs, local services, hardware)

Decision Point: You should be able to articulate:

  • "This app needs access to [X] system resources"
  • "The main user flows are [Y]"
  • "Security sensitivity level is [Z]"

Step 2: Architecture Design (Security First)

Goal: Design a secure Main/Renderer architecture.

Thinking Framework - Security Principles:

  1. Least Privilege: Renderer should have minimal capabilities
  2. Defense in Depth: Multiple layers of protection
  3. Explicit Communication: All IPC channels are explicit and validated

Architecture Decision Matrix:

Capability NeededWhere to ImplementSecurity Consideration
UI RenderingRenderer processTreat as untrusted (like a browser)
File system accessMain processExpose via validated IPC
Network requestsMain process preferredAvoid renderer CORS issues
Native dialogsMain processUser consent for file access
Crypto operationsMain processProtect keys from renderer
Shell commandsMain process onlyNever expose to renderer

Decision Point: For each feature, answer:

  • "Does this need Main process access?"
  • "What is the minimal IPC surface needed?"

Step 3: IPC Design

Goal: Design safe, type-safe IPC communication.

Thinking Framework:

  • "What data flows between Main and Renderer?"
  • "Who initiates the communication?"
  • "What validation is needed on each end?"

IPC Pattern Selection:

Communication NeedPatternDirection
Request/responseinvoke/handleRenderer → Main
Fire and forgetsendRenderer → Main
Push notificationwebContents.sendMain → Renderer
Two-way streamMessagePortBidirectional

IPC Security Checklist:

  • All channels have explicit names
  • Input validation on Main process handlers
  • No arbitrary code execution from renderer input
  • Sensitive operations require user confirmation
  • Rate limiting for expensive operations

Type Safety Pattern:

// Define channel types in shared/
interface IpcChannels {
  'file:open': { args: void; return: string | null };
  'file:save': { args: { path: string; content: string }; return: boolean };
}

Step 4: Preload Script Design

Goal: Create a minimal, secure bridge between worlds.

Thinking Framework:

  • "What is the absolute minimum the renderer needs?"
  • "Am I exposing more than necessary?"
  • "Is each exposed function validated?"

Preload Design Principles:

  1. Minimal Surface: Only expose what's absolutely needed
  2. No Raw IPC: Wrap ipcRenderer, don't expose directly
  3. Type Definitions: Provide TypeScript types for renderer
  4. One-Way Binding: Prefer invoke over send/on pairs

Anti-Patterns to Avoid:

// BAD: Exposes raw ipcRenderer
contextBridge.exposeInMainWorld('electron', { ipcRenderer });

// BAD: Arbitrary channel execution
contextBridge.exposeInMainWorld('api', {
  send: (channel, data) => ipcRenderer.send(channel, data)
});

// GOOD: Explicit, limited API
contextBridge.exposeInMainWorld('api', {
  openFile: () => ipcRenderer.invoke('dialog:openFile'),
  saveFile: (content: string) => ipcRenderer.invoke('file:save', content)
});

Step 5: Window Management Strategy

Goal: Design appropriate window management for the application.

Thinking Framework:

  • "How many windows does this app need?"
  • "How do windows communicate?"
  • "What happens when windows are closed?"

Window Patterns:

App TypePattern
Single documentOne main window
Multi-documentWindow per document, shared state in Main
Dashboard + detailsParent-child windows
System utilityTray app with popup

Window Configuration Checklist:

  • Appropriate webPreferences for each window type
  • Window state persistence (position, size)
  • Proper close/quit behavior (hide vs destroy)
  • Deep linking / protocol handling

Step 6: Data Persistence Strategy

Goal: Design secure, reliable data storage.

Thinking Framework:

  • "What data needs to persist?"
  • "How sensitive is this data?"
  • "Does data need to sync across devices?"

Storage Options:

Data TypeSolutionSecurity
User preferenceselectron-storePlain or encrypted
Structured dataSQLite (better-sqlite3)File-level encryption
Large filesFile systemOS-level permissions
Credentialssystem keychain (keytar)OS secure storage

Data Security Checklist:

  • Sensitive data encrypted at rest
  • Credentials in system keychain, not files
  • Backup/export functionality
  • Data migration strategy for updates

Step 7: Packaging and Distribution

Goal: Configure reliable cross-platform distribution.

Thinking Framework:

  • "Which platforms are targets?"
  • "How will updates be delivered?"
  • "What signing/notarization is needed?"

Platform Checklist:

PlatformSigningDistribution
macOSDeveloper ID + NotarizationDMG, PKG, or Mac App Store
WindowsCode signing certificateNSIS, MSI, or Microsoft Store
LinuxOptional GPGAppImage, deb, rpm, Snap

Auto-Update Strategy:

  • electron-updater configuration
  • Update server (GitHub releases, S3, etc.)
  • Staged rollouts for critical updates
  • Rollback capability

Step 8: Testing Strategy

Goal: Ensure the application is reliable across platforms.

Testing Layers:

  • Unit Tests: Business logic in Main process
  • Integration Tests: IPC communication
  • E2E Tests: Spectron/Playwright for UI flows
  • Platform Tests: CI matrix for all target platforms

Testing Checklist:

  • Test IPC handlers in isolation
  • Test preload script type contracts
  • E2E tests for critical user flows
  • Platform-specific behavior tests

Usage

Scaffold New Project

bash /mnt/skills/user/electron-architect/scripts/scaffold-project.sh [project-name] [ui-framework] [package-manager]

Arguments:

  • project-name - Name of the project (default: my-electron-app)
  • ui-framework - UI framework: vanilla, react, svelte, vue (default: vanilla)
  • package-manager - Package manager: pnpm, npm, yarn (default: pnpm)

Examples:

bash /mnt/skills/user/electron-architect/scripts/scaffold-project.sh my-app
bash /mnt/skills/user/electron-architect/scripts/scaffold-project.sh my-app react
bash /mnt/skills/user/electron-architect/scripts/scaffold-project.sh my-app svelte pnpm

Security defaults:

  • nodeIntegration: false
  • contextIsolation: true
  • sandbox: true

Documentation Resources

Official Documentation:

  • Electron: https://www.electronjs.org/docs/latest/
  • Electron Forge: https://www.electronforge.io/
  • electron-builder: https://www.electron.build/

Project Structure

src/
├── main/
│   ├── main.ts              # Main process entry
│   ├── ipc/                  # IPC handlers
│   │   └── file-handlers.ts
│   ├── services/             # Backend services
│   │   └── database.ts
│   └── menu.ts               # Application menu
├── preload/
│   └── preload.ts            # Context bridge
├── renderer/                 # UI (React/Svelte/Vue)
│   ├── App.tsx
│   └── components/
└── shared/
    └── types.ts              # Shared type definitions

Security Configuration

BrowserWindow Settings

// main.ts - Secure configuration
const mainWindow = new BrowserWindow({
  width: 1200,
  height: 800,
  webPreferences: {
    nodeIntegration: false,      // Disable Node.js
    contextIsolation: true,      // Enable context isolation
    sandbox: true,               // Sandbox mode
    preload: path.join(__dirname, 'preload.js'),
    webSecurity: true,           // Enforce same-origin
  }
});

Preload Script Pattern

// preload.ts - Safe API exposure
import { contextBridge, ipcRenderer } from 'electron';

contextBridge.exposeInMainWorld('electronAPI', {
  // One-way: Renderer → Main
  saveFile: (content: string) =>
    ipcRenderer.invoke('file:save', content),

  // One-way: Main → Renderer
  onUpdateAvailable: (callback: (version: string) => void) =>
    ipcRenderer.on('update-available', (_, version) => callback(version)),

  // Request-response pattern
  openFile: () => ipcRenderer.invoke('dialog:openFile'),
});

// Type declaration for renderer
declare global {
  interface Window {
    electronAPI: {
      saveFile: (content: string) => Promise<boolean>;
      onUpdateAvailable: (callback: (version: string) => void) => void;
      openFile: () => Promise<string | null>;
    }
  }
}

Main Process Handlers

// main/ipc/file-handlers.ts
import { ipcMain, dialog } from 'electron';
import { readFile, writeFile } from 'fs/promises';

export function registerFileHandlers() {
  ipcMain.handle('dialog:openFile', async () => {
    const { canceled, filePaths } = await dialog.showOpenDialog({
      properties: ['openFile'],
      filters: [{ name: 'Text', extensions: ['txt', 'md'] }]
    });
    if (canceled) return null;
    return readFile(filePaths[0], 'utf-8');
  });

  ipcMain.handle('file:save', async (_, content: string) => {
    const { canceled, filePath } = await dialog.showSaveDialog({});
    if (canceled || !filePath) return false;
    await writeFile(filePath, content);
    return true;
  });
}

IPC Communication Patterns

Pattern 1: Invoke (Request-Response)

// Renderer
const data = await window.electronAPI.fetchData(id);

// Main
ipcMain.handle('fetch-data', async (event, id) => {
  return await database.get(id);
});

Pattern 2: Send/On (Fire-and-Forget)

// Main → Renderer
mainWindow.webContents.send('notification', message);

// Renderer
window.electronAPI.onNotification((msg) => showToast(msg));

Pattern 3: Two-Way Events

// Renderer sends, awaits Main response
const result = await window.electronAPI.processFile(path);

Packaging Configuration

Electron Forge

{
  "config": {
    "forge": {
      "packagerConfig": {
        "asar": true,
        "icon": "./assets/icon"
      },
      "makers": [
        { "name": "@electron-forge/maker-squirrel" },
        { "name": "@electron-forge/maker-dmg" },
        { "name": "@electron-forge/maker-deb" }
      ]
    }
  }
}

Present Results to User

When providing Electron solutions:

  • Always follow security best practices
  • Provide complete IPC communication examples
  • Consider cross-platform compatibility
  • Include TypeScript types for the API
  • Note Electron version differences

Troubleshooting

"require is not defined"

  • nodeIntegration is correctly disabled
  • Use preload script with contextBridge

"Cannot access window.electronAPI"

  • Check preload script path is correct
  • Verify contextIsolation is true
  • Ensure contextBridge.exposeInMainWorld is called

"IPC message not received"

  • Verify channel names match exactly
  • Check if handler is registered before window loads
  • Use invoke for async responses

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

能力 5

展示第三方安全扫描或审计结果

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

29.91%
按下载量换算223

OpenCode

24.34%
按下载量换算181

Cursor

18.16%
按下载量换算135

Antigravity

12.62%
按下载量换算94

github-copilot

7.66%
按下载量换算57

Codex

3.57%
按下载量换算27

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

操作浏览器

该 Skill 可能涉及浏览器控制能力,使用时可能读取或操作网页内容,需要在受控环境中确认权限边界。

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。

来源信息

继续浏览同类 Skills