shellx-ai 1.0.8 → 1.0.9

package/dist/types.d.ts

@@ -0,0 +1,127 @@
+export type Point = {
+    x: number;
+    y: number;
+};
+export type BaseOptions = {
+    wait?: number;
+    timeout?: number;
+    retry?: number;
+    waitAfterMs?: number;
+};
+export type Click = BaseOptions & {
+    targetElementId?: string;
+    targetResourceId?: string;
+    targetText?: string;
+    targetClass?: string;
+    targetX?: number;
+    targetY?: number;
+    clickType?: 'single' | 'long' | 'double';
+    visible?: boolean;
+    clickable?: boolean;
+};
+export type Input = BaseOptions & {
+    text: string;
+    targetElementId?: string;
+    targetResourceId?: string;
+    targetText?: string;
+    targetClass?: string;
+    clear?: boolean;
+    hideKeyboard?: boolean;
+    visible?: boolean;
+    clickable?: boolean;
+};
+export type Swipe = BaseOptions & {
+    fromX: number;
+    fromY: number;
+    toX: number;
+    toY: number;
+    duration?: number;
+};
+export type Key = BaseOptions & {
+    key: string;
+    longPress?: boolean;
+};
+export type Wait = BaseOptions & {
+    targetElementId?: string;
+    targetResourceId?: string;
+    targetText?: string;
+    targetClass?: string;
+    condition?: 'visible' | 'clickable' | 'gone';
+    visible?: boolean;
+    clickable?: boolean;
+};
+export type Find = BaseOptions & {
+    targetElementId?: string;
+    targetResourceId?: string;
+    targetText?: string;
+    targetClass?: string;
+    multiple?: boolean;
+    maxResults?: number;
+    visible?: boolean;
+    clickable?: boolean;
+    pressClick?: boolean;
+};
+export type Command = BaseOptions & {
+    cmd: string;
+};
+export type AppInfo = BaseOptions & {
+    package: string;
+};
+export type Screenshot = BaseOptions & {
+    regionX?: number;
+    regionY?: number;
+    regionWidth?: number;
+    regionHeight?: number;
+    format?: 'png' | 'jpeg';
+    quality?: number;
+};
+export type ActionResult = {
+    success: boolean;
+    data?: any;
+    error?: string;
+    duration: number;
+};
+export type Element = {
+    id: string;
+    text: string;
+    class: string;
+    left: number;
+    top: number;
+    right: number;
+    bottom: number;
+    visible: boolean;
+    clickable: boolean;
+};
+export type FindResult = {
+    elements: Element[];
+    count: number;
+    success: boolean;
+    found: boolean;
+};
+export type CommandResult = {
+    success: boolean;
+    output: string;
+    error?: string;
+    exitCode?: number;
+    duration: number;
+};
+export type Action = ({
+    type: 'click';
+} & Click) | ({
+    type: 'input';
+} & Input) | ({
+    type: 'swipe';
+} & Swipe) | ({
+    type: 'key';
+} & Key) | ({
+    type: 'wait';
+} & Wait) | ({
+    type: 'find';
+} & Find) | ({
+    type: 'command';
+} & Command) | ({
+    type: 'appInfo';
+} & AppInfo) | ({
+    type: 'screenshot';
+} & Screenshot);
+export type Actions = Action[];

package/dist/types.js

@@ -0,0 +1,4 @@
+"use strict";
+// 精简的 ShellX 类型定义
+// 为了让大模型更容易理解参数，所有类型都设计为最简洁的扁平结构
+Object.defineProperty(exports, "__esModule", { value: true });

package/dist/utils.d.ts

@@ -11,8 +11,3 @@ export declare function setEnvVar(key: string, value: string): void;
  * 安全地删除环境变量，兼容浏览器和Node.js环境
  */
 export declare function deleteEnvVar(key: string): void;
-/**
- * 检查并获取 WebSocket URL 环境变量
- * 如果未设置则显示错误并抛出异常
- */
-export declare function getWebSocketUrl(): string;

package/dist/utils.js

@@ -4,7 +4,6 @@ exports.wait = void 0;
 exports.getEnvVar = getEnvVar;
 exports.setEnvVar = setEnvVar;
 exports.deleteEnvVar = deleteEnvVar;
-exports.getWebSocketUrl = getWebSocketUrl;
 const wait = (ms) => new Promise(resolve => setTimeout(resolve, ms));
 exports.wait = wait;
 /**
@@ -47,35 +46,3 @@ function deleteEnvVar(key) {
         console.log("已删除环境变量:", key);
     }
 }
-/**
- * 检查并获取 WebSocket URL 环境变量
- * 如果未设置则显示错误并抛出异常
- */
-function getWebSocketUrl() {
-    // 安全地获取环境变量，兼容浏览器和Node.js环境
-    const wsUrl = getEnvVar('WEBSOCKET_URL');
-    if (!wsUrl || wsUrl.trim() === '') {
-        const errorMessage = '❌ 错误: 未设置 WEBSOCKET_URL 环境变量！\n\n📋 解决方案：\n1. 在项目根目录创建 .env 文件\n2. 在 .env 文件中添加以下内容：\n   WEBSOCKET_URL=ws://127.0.0.1:9091/api/s/your-session-id\n\n💡 提示：\n- 请将 your-session-id 替换为实际的会话ID\n- 确保 ShellX 服务器正在运行\n- 检查端口号是否正确（默认9091）';
-        console.error(errorMessage);
-        // 在Node.js环境中退出，在浏览器环境中抛出异常
-        if (typeof process !== 'undefined' && process.exit) {
-            process.exit(1);
-        }
-        else {
-            throw new Error(errorMessage);
-        }
-    }
-    // 简单的 URL 格式验证
-    if (!wsUrl.startsWith('ws://') && !wsUrl.startsWith('wss://')) {
-        const errorMessage = `❌ 错误: WEBSOCKET_URL 格式不正确！\n当前值: ${wsUrl}\n应该以 ws:// 或 wss:// 开头\n例如: ws://127.0.0.1:9091/api/s/your-session-id`;
-        console.error(errorMessage);
-        // 在Node.js环境中退出，在浏览器环境中抛出异常
-        if (typeof process !== 'undefined' && process.exit) {
-            process.exit(1);
-        }
-        else {
-            throw new Error(errorMessage);
-        }
-    }
-    return wsUrl;
-}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "shellx-ai",
-  "version": "1.0.8",
+  "version": "1.0.9",
   "description": "shellx is a powerful WebSocket-based client for controlling shell commands and UI automation on remote devices.",
   "repository": {
     "url": "git+https://github.com/10cl/shellx.git",
@@ -33,6 +33,7 @@
     "cbor": "^9.0.0",
     "dotenv": "^16.4.5",
     "ofetch": "^1.4.1",
+    "shellx-ai": "file:",
     "uuid": "^11.1.0"
   },
   "optionalDependencies": {

package/README.md

@@ -1,432 +0,0 @@
-# ShellX
-
-[![npm version](https://badge.fury.io/js/shellx.svg)](https://badge.fury.io/js/shellx)
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
-
-**ShellX** is a powerful automation framework for Android device control and UI automation. It provides seamless shell command execution, element finding, screen capture, and automated task execution through an intelligent ShellX.ai service architecture.
-
-## ✨ Features
-
-- 🔐 **Zero-Configuration Setup** - Automatic ShellX.ai authentication and service connection
-- 📱 **Android Device Control** - Execute shell commands with real-time output monitoring
-- 🎯 **Smart UI Automation** - Advanced element finding with retry logic and multiple strategies
-- 📸 **Screen Operations** - Screenshots, screen info, and visual change detection
-- 🔄 **Intelligent Fallback** - Automatic switching between ShellX.ai cloud and local services
-- 🛡️ **Robust Error Handling** - Comprehensive error recovery and diagnostics
-- 🧠 **Shell Output Monitoring** - Real-time command output processing and analysis
-- 📊 **Rich Diagnostics** - Built-in element analysis and debugging tools
-- 🔗 **Smart Connection Management** - ShellX.ai service connection status with message-based verification
-
-## 🚀 Quick Start
-
-### Installation
-
-```bash
-npm install shellx-ai
-```
-
-### Node.js Version Compatibility
-
-ShellX supports Node.js 14+ with automatic polyfills:
-
-**HTTP Requests:**
-- **Fallback**: Node.js 18+ built-in `fetch` or `node-fetch` for older versions
-- **Browser**: Compatible with all fetch implementations
-
-**ShellX.ai Service Connection:**
-- **Browser**: Uses native `WebSocket` API for service communication
-- **Node.js**: Automatically uses `ws` module for service communication
-
-No additional configuration needed - ShellX handles environment detection automatically!
-
-### Basic Setup
-
-```typescript
-import { createShellXWithShellMonitoring } from 'shellx';
-
-// Set your authentication key
-process.env.SHELLX_AUTH_KEY = 'your-auth-key';
-
-// Initialize ShellX with automatic ShellX.ai authentication
-const { client, shellx } = await createShellXWithShellMonitoring();
-
-// Start automating!
-await shellx.executeShellCommand('getprop ro.build.version.release');
-await shellx.clickByText('Settings');
-```
-
-### Environment Setup
-
-Set your authentication key as an environment variable:
-
-```bash
-# Linux/macOS
-export SHELLX_AUTH_KEY="your-auth-key"
-
-# Windows
-set SHELLX_AUTH_KEY=your-auth-key
-```
-
-## 📱 Common Use Cases
-
-### 1. Shell Command Execution
-
-```typescript
-// Execute shell commands with real-time output
-const result = await shellx.executeShellCommand('pm list packages', {
-  title: 'List installed packages',
-  timeout: 10000,
-  allowPartialResult: true, // Return partial results on timeout instead of throwing error
-  onOutput: (output) => {
-    console.log('Real-time output:', output);
-  }
-});
-
-console.log('Command result:', result.output);
-console.log('Success:', result.success);
-console.log('Duration:', result.duration, 'ms');
-```
-
-### 2. UI Element Operations
-
-```typescript
-// Smart element finding with retry logic
-const elements = await shellx.findElementsWithRetry({
-  className: 'android.widget.TextView',
-  textContains: 'Settings'
-}, 3, 1000, {
-  maxResults: 10,
-  visibleOnly: true
-});
-
-// Print detailed element information
-shellx.printElementsInfo(elements, 'Found Settings Elements');
-
-// Click by text content
-await shellx.clickByText('Wi-Fi', false);
-
-// Input text with options
-await shellx.inputText({
-  resourceId: 'search_field'
-}, 'Hello ShellX', {
-  clear: true,
-  hideKeyboard: true
-});
-```
-
-### 3. Screen Operations
-
-```typescript
-// Capture screenshot with details
-const screenshot = await shellx.captureScreen({
-  format: 'png',
-  quality: 90,
-  saveInfo: true
-});
-
-// Swipe gestures
-await shellx.swipe('up', 400, 800);  // direction, distance, duration
-await shellx.swipe('left', 300, 600);
-```
-
-### 4. App Automation Workflow
-
-```typescript
-// Login automation example
-await shellx.performLogin(
-  { resourceId: 'username_field' },      // username selector
-  { resourceId: 'password_field' },      // password selector  
-  { resourceId: 'login_button' },        // login button selector
-  'user@example.com',                    // username
-  'password123'                          // password
-);
-
-// Navigation workflow
-await shellx.navigateByPath([
-  'Settings',
-  'Network & internet', 
-  'Wi-Fi'
-]);
-```
-
-### 5. Advanced Element Finding
-
-```typescript
-// Wait for any of multiple elements to appear
-const result = await shellx.waitForAnyElement([
-  { textContains: 'OK' },
-  { textContains: 'Cancel' },
-  { textContains: 'Skip' }
-], 10000);
-
-if (result) {
-  console.log(`Found element: ${result.element.text} at index ${result.selectorIndex}`);
-}
-
-// Scroll to find element
-const element = await shellx.scrollToFindElement({
-  textContains: 'Privacy Settings'
-}, 5, 'down');
-```
-
-### 6. Batch Shell Commands
-
-```typescript
-// Execute multiple commands in sequence
-const commands = [
-  { command: 'am force-stop com.example.app', title: 'Stop app' },
-  { command: 'am start -n com.example.app/.MainActivity', title: 'Start app' },
-  { command: 'input keyevent KEYCODE_HOME', title: 'Go home', waitAfterMs: 1000 }
-];
-
-const results = await shellx.executeShellCommands(commands, {
-  continueOnError: true,
-  timeout: 5000
-});
-```
-
-### 7. Device Information
-
-```typescript
-// Get comprehensive device information
-const deviceInfo = await shellx.getDeviceInfo();
-deviceInfo.forEach((info, index) => {
-  console.log(`${index + 1}. ${info.output}`);
-});
-
-// ADB commands with enhanced output
-const result = await shellx.adbCommand('shell dumpsys battery', {
-  title: 'Get battery status',
-  successPattern: /level: \d+/,
-  errorPattern: /error|failed/i
-});
-```
-
-
-
-## 🎯 ShellX Class API
-
-### Core Methods
-
-| Method | Description | Example |
-|--------|-------------|---------|
-| `executeShellCommand(cmd, opts?)` | Execute shell command with monitoring | `shellx.executeShellCommand('ls -la', {allowPartialResult: true})` |
-| `completeTask(taskId, result?, success?)` | Manually complete a specific task | `client.completeTask('task-123', {output: 'result'})` |
-| `getPendingTaskIds()` | Get all pending task IDs | `client.getPendingTaskIds()` |
-| `getTaskInfo(taskId)` | Get task information | `client.getTaskInfo('task-123')` |
-| `completeTasksByType(type, result?, success?)` | Complete all tasks of a specific type | `client.completeTasksByType('action')` |
-| `clickByText(text, exact?)` | Click element by text content | `shellx.clickByText('OK', true)` |
-| `inputText(selector, text, opts?)` | Input text into element | `shellx.inputText({resourceId: 'field'}, 'text')` |
-| `findElementWithRetry(selector, retries?, delay?)` | Find single element with retry | `shellx.findElementWithRetry({text: 'Button'})` |
-| `findElementsWithRetry(selector, retries?, delay?, opts?)` | Find multiple elements with retry | `shellx.findElementsWithRetry({className: 'TextView'})` |
-| `swipe(direction, distance?, duration?)` | Perform swipe gesture | `shellx.swipe('up', 400, 800)` |
-| `captureScreen(opts?)` | Take screenshot | `shellx.captureScreen({saveInfo: true})` |
-
-### Utility Methods
-
-| Method | Description |
-|--------|-------------|
-| `printElementInfo(element, index?)` | Print detailed element information |
-| `printElementsInfo(elements, title?)` | Print information for multiple elements |
-| `waitForAnyElement(selectors, timeout?)` | Wait for any of multiple elements |
-| `scrollToFindElement(selector, maxScrolls?, direction?)` | Scroll to find element |
-| `performLogin(userSel, passSel, btnSel, user, pass)` | Automated login flow |
-| `navigateByPath(textPath)` | Navigate through UI by text path |
-| `getDeviceInfo()` | Get comprehensive device information |
-
-## 🛠️ Configuration
-
-### Environment Variables
-
-```bash
-# Required: Your ShellX authentication key
-SHELLX_AUTH_KEY=your-auth-key
-
-# Optional: Timeout settings
-SHELLX_TIMEOUT=20000
-SHELLX_RETRY_ATTEMPTS=3
-```
-
-### Connection Options
-
-```typescript
-const { client, shellx } = await createShellXWithShellMonitoring({
-  timeout: 15000,
-  reconnectMaxAttempts: 3,
-  onOpen: () => console.log('Connected'),
-  onClose: (event) => console.log('Disconnected'),
-  onError: (error) => console.error('Error:', error),
-  onMessage: (data) => {
-    // Handle custom messages
-  }
-});
-```
-
-## 🔧 Advanced Features
-
-### Shell Command Options
-
-ShellX provides flexible options for shell command execution:
-
-```typescript
-const result = await shellx.executeShellCommand('long-running-command', {
-  title: 'Custom command title',
-  timeout: 30000,                    // Command timeout in milliseconds
-  allowPartialResult: true,          // Return partial results on timeout instead of throwing error
-  waitAfterMs: 1000,                 // Wait time after command completion
-  onOutput: (output) => {            // Real-time output callback
-    console.log('Live output:', output);
-  },
-  onError: (error) => {              // Error callback
-    console.error('Command error:', error);
-  },
-  successPattern: /success|completed/, // Pattern to detect success
-  errorPattern: /error|failed/        // Pattern to detect failure
-});
-```
-
-#### allowPartialResult Parameter
-
-The `allowPartialResult` parameter controls timeout behavior:
-
-- **`true`**: When command times out but has partial output, returns the partial result instead of throwing an error
-- **`false`**: When command times out, always throws an error (default behavior)
-- **`undefined`**: Uses default behavior (returns partial results if available)
-
-```typescript
-// Example: Allow partial results for long-running commands
-const result = await shellx.executeShellCommand('find / -name "*.log"', {
-  timeout: 5000,
-  allowPartialResult: true
-});
-
-if (!result.success && result.error?.includes('timeout')) {
-  console.log('Command timed out, but got partial results:', result.output);
-} else {
-  console.log('Command completed successfully:', result.output);
-}
-```
-
-### Task Management
-
-ShellX provides advanced task management capabilities for manual control:
-
-```typescript
-// Send message and get taskId for manual control
-const { taskId, promise } = await client.sendMessageWithTaskId(
-  { actions: { type: 'command', command: 'long-running-command' } },
-  'action',
-  'command'
-);
-
-console.log('Task ID:', taskId);
-
-// Manually complete the task when ready
-client.completeTask(taskId, { output: 'Command completed' }, true);
-
-// Get all pending task IDs
-const pendingTasks = client.getPendingTaskIds();
-console.log('Pending tasks:', pendingTasks);
-
-// Get task information
-const taskInfo = client.getTaskInfo(taskId);
-console.log('Task info:', taskInfo);
-
-// Complete all tasks of a specific type
-const completedCount = client.completeTasksByType('action', { success: true });
-console.log(`Completed ${completedCount} action tasks`);
-```
-
-### Shell Output Monitoring
-
-ShellX automatically monitors shell command output in real-time:
-
-```typescript
-const result = await shellx.executeShellCommand('top -n 1', {
-  onOutput: (output) => {
-    console.log('Live output:', output);
-  },
-  successPattern: /Tasks:/,
-  errorPattern: /error|failed/i,
-  timeout: 10000,
-  allowPartialResult: true // Return partial results on timeout
-});
-```
-
-
-
-### Element Diagnostics
-
-Built-in diagnostic tools for troubleshooting:
-
-```typescript
-// Automatic element diagnosis when finding fails
-const elements = await shellx.findElementsWithRetry({
-  textContains: 'NotFound'
-});
-
-if (elements.length === 0) {
-  // ShellX automatically runs diagnostics
-  // Shows available elements, device info, etc.
-}
-```
-
-### Intelligent Fallback
-
-ShellX automatically handles service availability:
-
-```typescript
-// 1. Tries online authentication: https://shellx.ai/api/device/{authKey}
-// 2. On success: Uses returned machine URL  
-// 3. On failure: Falls back to local: ws://127.0.0.1:9091/api/s/{authKey}
-// 4. Seamless operation regardless of network conditions
-```
-
-## 📚 Complete Examples
-
-Check out the [examples directory](./examples/) for comprehensive use cases:
-
-- **[basic.ts](./examples/basic.ts)** - Basic automation operations
-- **[shell-commands-demo.ts](./examples/shell-commands-demo.ts)** - Shell command execution
-- **[find-elements-demo.ts](./examples/find-elements-demo.ts)** - Element finding strategies
-
-## 🚨 Error Handling
-
-ShellX provides comprehensive error handling with automatic recovery:
-
-```typescript
-try {
-  await shellx.executeShellCommand('invalid-command');
-} catch (error) {
-  // ShellX handles timeouts, connection issues, and command failures
-  console.error('Command failed:', error.message);
-  
-  // Automatic diagnostics on failure
-  // Device info, available elements, etc.
-}
-```
-
-## 🤝 Contributing
-
-Contributions are welcome! Please feel free to submit a Pull Request.
-
-## 📄 License
-
-MIT License - see the [LICENSE](LICENSE) file for details.
-
-## 🆘 Support
-
-- 📚 [Documentation](https://shellx.ai/docs)
-- 🐛 [GitHub Issues](https://github.com/your-org/shellx/issues)
-- 💬 [Discord Community](https://discord.gg/shellx)
-
-## 🎉 Getting Started
-
-1. **Install ShellX**: `npm install shellx-ai`
-2. **Set your auth key**: `export SHELLX_AUTH_KEY="your-key"`  
-3. **Run the example**: `ts-node examples/basic.ts`
-4. **Start automating!** 🚀
-
----
-
-*Made with ❤️ for the automation community* 

package/dist/constants.d.ts

@@ -1,13 +0,0 @@
-export interface TaskClientConfig {
-    timeout: number;
-    reconnect: boolean;
-    reconnectMaxAttempts: number;
-    reconnectInterval: number;
-    pingInterval: number;
-    onOpen: () => void;
-    onClose: (event?: CloseEvent) => void;
-    onError: (error?: Event) => void;
-    onReconnectFailed: () => void;
-    onMessage?: (data: any) => void;
-}
-export declare const DEFAULT_CONFIG: TaskClientConfig;

package/dist/constants.js

@@ -1,14 +0,0 @@
-"use strict";
-Object.defineProperty(exports, "__esModule", { value: true });
-exports.DEFAULT_CONFIG = void 0;
-exports.DEFAULT_CONFIG = {
-    timeout: 5000,
-    reconnect: true,
-    reconnectMaxAttempts: 5,
-    reconnectInterval: 1000,
-    pingInterval: 2000,
-    onOpen: () => { },
-    onClose: () => { },
-    onError: () => { },
-    onReconnectFailed: () => { },
-};