devicely 2.3.3 → 2.3.5

package/README.md

@@ -1,321 +1,240 @@
 <div align="center">
 
-# 📱✨ Devicely
+# Devicely
 
-### **Control one device, and every connected device follows in real-time - with live screen sync on the web**
+### Tap once. Every iPhone, iPad and Android follows.
 
-*Ever wondered what happens when you combine device mirroring with AI-powered automation? Meet the mobile orchestration platform that curious developers are exploring.*
+Control a fleet of real iOS and Android devices from one screen — by hand, by voice, or through AI agents over MCP. Runs locally on your machine, against your own devices.
 
-### 🤔 **Intrigued? Try it yourself:**
+[![npm version](https://img.shields.io/npm/v/devicely?style=for-the-badge&logo=npm&color=CB3837)](https://www.npmjs.com/package/devicely)
+[![npm downloads](https://img.shields.io/npm/dm/devicely?style=for-the-badge&color=CB3837)](https://www.npmjs.com/package/devicely)
+[![Platforms](https://img.shields.io/badge/devices-iOS%20%7C%20iPadOS%20%7C%20Android-lightgrey?style=for-the-badge)](#supported-platforms)
+[![MCP](https://img.shields.io/badge/MCP-30%20tools-8A2BE2?style=for-the-badge)](#for-ai-agents-mcp-server)
+[![License](https://img.shields.io/badge/license-free%20%7C%20closed%20source-blue?style=for-the-badge)](#license)
+
+<img src="https://raw.githubusercontent.com/sureshkumarm8/Devicely/main/DevicelyDemo.gif" alt="One tap on the commander device is mirrored on every follower device" width="900">
+
+[Quick start](#quick-start) · [For QA teams](#for-qa-teams) · [For AI agents](#for-ai-agents-mcp-server) · [CLI](#cli-reference) · [Supported platforms](#supported-platforms) · [Privacy](#privacy-and-security)
+
+</div>
+
+---
+
+## Quick start
 
 ```bash
 npm install -g devicely
-devicely start
+devicely doctor      # checks Node, adb, Xcode and device drivers
+devicely start       # starts Devicely and opens the Command Center
 ```
-**Visit [https://devicely-ai.vercel.app/](https://devicely-ai.vercel.app/) and see what happens when devices work together.**
 
-[![NPM Version](https://img.shields.io/badge/npm-v2.3.0-CB3837?style=for-the-badge)](https://www.npmjs.com/package/devicely)
-[![License](https://img.shields.io/badge/license-MIT-green?style=for-the-badge)](LICENSE)
-[![Platform](https://img.shields.io/badge/platform-iOS%20%7C%20Android-lightgrey?style=for-the-badge)](https://devicely-ai.vercel.app)
+The Command Center opens at **[devicely-ai.vercel.app](https://devicely-ai.vercel.app)**. Plug in a device over USB, or click **Connect via QR** to pair one over Wi-Fi.
 
-[🚀 Quick Start](#-quick-start) • [✨ Key Features](#-the-devicely-advantage) • [🌌 Infinite Possibilities](#-glimpse-of-the-infinite) • [📟 CLI Reference](#-cli-reference)
+**No phone handy?** Boot an Android emulator or iOS simulator with the server:
 
----
+```bash
+devicely emulators list            # shows installed AVDs and simulators
+devicely start --launch <name>     # starts Devicely and boots that virtual device
+```
 
-### 🤷‍♀️ **What if mobile testing was more like conducting an orchestra?**
+**Requirements:** Node.js 18+ on macOS or Linux. Android needs `adb` (platform-tools) on your `PATH`. iOS and iPadOS devices need a Mac with Xcode — see [iOS setup](#ios-setup).
 
-Instead of manually testing the same flow across multiple devices, what if you could:
+---
 
-- **Speak naturally** to your devices: *"Take a screenshot"* → all devices respond
-- **Touch one screen** and watch others mirror your actions in real-time  
-- **Record once**, replay everywhere - from 1 device to 100+
-- **See everything** happening across your device fleet on one web dashboard
+## Why Devicely
 
-*Turns out, when devices can follow a leader and understand natural language, mobile development gets pretty interesting.*
+Testing one flow on ten devices usually means doing it ten times by hand, renting cloud devices by the minute, or maintaining per-device scripts. Devicely takes a different approach: **you drive one device, and the rest of your fleet does the same thing at the same moment.**
 
-</div>
+| | Devicely | Cloud device labs | Script frameworks (Appium, Maestro, Detox) | Screen mirroring (scrcpy, Vysor) |
+| :--- | :---: | :---: | :---: | :---: |
+| Live 1-to-many gestures (one tap → every device) | ✅ | ❌ | ❌ | ❌ |
+| iOS and Android mixed in one live session | ✅ | ❌ | ❌ | ❌ |
+| Uses your own devices, on your own machine | ✅ | ❌ | ✅ | ✅ |
+| No code needed (UI, voice, plain English) | ✅ | Partial | ❌ | ✅ (one device) |
+| Built-in MCP server for AI agents | ✅ | Varies | Varies | ❌ |
+| Exports locators to test code | ✅ | — | ✅ | ❌ |
 
----
+Devicely doesn't replace your test framework. It shortens the slow parts around it: exploring, reproducing bugs across devices, smoke-testing builds, and finding the locators you'll put in your tests.
 
 ---
 
-## 🌟 **What makes this different?**
-
-### 👑 **Commander Mode** - *Like screen sharing, but for device control*
-*   One device leads, others follow in real-time
-*   Touch the leader's screen → all followers mirror the action
-*   Switch leaders with a simple dropdown
-*   Watch it all happen on your web dashboard
-*   **[NEW in v2.3.0]** Enhanced WDA API for iOS device control
-*   **[NEW in v2.3.0]** Element caching for faster command execution
-
-### 📱 **QR-Based Device Enrollment** - *Connect devices in seconds, not minutes*
-*   **[NEW in v2.3.0]** Scan QR code to enroll iOS or Android devices
-*   **[NEW in v2.3.0]** Automatic IP address detection and configuration
-*   **[NEW in v2.3.0]** Works with both USB and wireless connections
-*   iOS: WebDriverAgent auto-detection
-*   Android: Automatic ADB setup with privacy controls
-*   No more manual IP configuration or device.conf editing
-
-### 🔄 **Live Sync** - *What if all your devices moved together?*
-*   Commands reach all devices simultaneously via WebSockets
-*   iOS and Android understand the same instructions differently (as they should)
-*   See live status updates from every connected device
-*   No waiting for one device to finish before the next starts
-*   **[NEW in v2.3.0]** Improved wireless connectivity with retry logic
-
-### 📹 **Record & Replay - Zero-Code Automation**
-*   **Visual Recording:** Capture complex workflows with simple button clicks
-*   **AI-Generated Scripts:** Voice commands auto-convert to replayable recordings
-*   **Fleet Replay:** Run recorded flows on single devices or entire laboratories
-*   **Edit & Refine:** Full editor for recordings with add/remove/modify capabilities
-
-### 🤖 **AI Integration** - *Talk to your devices like a person*
-*   Say *"Open Instagram and find search"* - it just works
-*   Choose from 7 AI providers with latest models:
-    - **OpenAI:** GPT-4o (Aug 2024)
-    - **Google:** Gemini 1.5 Pro/Flash
-    - **Anthropic:** Claude 3.5 Sonnet (Oct 2024)
-    - **Groq:** Llama 3.3 70B (ultra-fast LPU inference)
-    - **Cohere:** Command R+ (enterprise)
-    - **Mistral:** Mistral Large 2
-    - **GitHub:** Copilot powered
-*   **[NEW in v2.3.0]** API-only mode - no local models, optimized for speed
-*   Devices understand context - no hunting for element IDs
-*   Same command, different results on iOS vs Android (intelligently)
-
-### 🎤 **Voice Control** - *Because typing is so 2020*
-*   Speak naturally, pause when you're thinking (2-second detection)
-*   Chain commands: *"launch settings, scroll up, launch camera"*
-*   Watch your words turn into actions in real-time
-*   Works across as many devices as you want to connect
+## For QA teams
 
----
+### Commander Mode — one-to-many control
 
-## ✨ **Other things you might find useful**
+- **Live mirror of the commander device**, with every follower shown alongside it.
+- **Every gesture is broadcast**: tap, long-press, swipe (speed preserved) and typed text go to all followers at once.
+- **Element-aware, not pixel-blind.** Followers find the element the commander touched — by accessibility ID, label, type and position in the UI tree — so a tap on an iPhone hits the same button on an iPad or a Pixel, even when the layouts differ.
+- **Swap the commander** from a dropdown on the mirror, without restarting the session.
+- **Include or exclude individual followers** with one click.
+- **Fast iOS mirroring** over WebDriverAgent MJPEG: about 170 KB per frame, compared with about 5.8 MB for a screenshot poll. Each iOS device gets its own WDA port, so several iPhones don't collide.
+- **Full-screen mirror** with zoom and fit-to-window, and a **D-pad** for directional scrolling and swiping.
 
-### ⚡ **Quick Actions Panel**
-*   **Type Text to Send:** Rapid text input without typing full commands
-*   **Element to Tap:** Click by name or coordinates (x,y) instantly
-*   **App ID to Launch:** Launch apps with autocomplete dropdown
-*   **History Navigation:** Arrow keys cycle through previous inputs
+### Recordings that survive UI changes
 
-### 🎮 **Unified Command Center**
-*   **Cross-Platform:** Native support for **iOS 14+** and **Android 5+**
-*   **Massive Scale:** Control 5, 50, or 100+ devices with zero overhead
-*   **Live Feedback:** Real-time logs and visual confirmations via WebSockets
-*   **API Key Management:** Secure local storage with show/hide toggles
+- **Steps are recorded as locators** (`accessibilityId`, `resourceId`, text, type, bounds), not raw coordinates.
+- **Auto-wait**: on replay, each step waits up to 10 s for its element to appear, so animations and slow networks don't break the run.
+- **Coordinate fallback**: if an element can't be found, Devicely taps its scaled position instead and marks the step `fallback`, so you can see what drifted.
+- **Replay on one device or the whole fleet in parallel.**
+- **Manage recordings** in a list or grid view with search and tag filters. A details drawer shows each step's locator, timing and result; the editor lets you add, edit, reorder and delete steps.
 
-### 🔍 **Smart UI Inspector**
-*   **Visual Element Scan:** See all clickable, enabled, visible elements instantly
-*   **Point-and-Click Control:** Direct device interaction through inspector
-*   **Developer Export:** Clean selectors and coordinates for automation scripts
+### Plain English and voice
 
-### 🌌 **The bigger picture**
-*   Transform 2-hour manual testing into 5-minute parallel runs
-*   Turn your device collection into a web-accessible lab
-*   Reproduce bugs instantly across different devices/OS versions
-*   Mix iOS and Android devices - they all speak the same language
+Type or say *"launch settings, scroll down, open camera"*. Devicely turns it into the right action for each platform — for example `com.apple.Preferences` on iOS and `com.android.settings` on Android — and runs it on every selected device at once. Voice input waits for a 2-second pause before it runs, so you can speak naturally.
 
----
+Choose your AI provider in **Settings → AI Config**:
 
-## 🔍 **How it works** *(in about 60 seconds)*
+| Provider | Notes |
+| :--- | :--- |
+| **Ollama** | Runs locally. No API key, and no prompt leaves your machine. |
+| **Google Gemini**, **Groq**, **Cohere** | Offer free tiers — the easiest way to start. |
+| **OpenAI**, **Anthropic Claude**, **Mistral AI** | Bring your own API key. |
+| **GitHub Copilot** | Uses your Copilot subscription. |
 
-### **Step 1: Get it running**
-```bash
-npm install -g devicely
-devicely start
-```
+Each provider lists its current models in the settings screen.
+
+### Quick actions, inspector and code export
 
-### **Step 2: Connect and explore**
-1. Open [https://devicely-ai.vercel.app/](https://devicely-ai.vercel.app/)
-2. Connect your iOS/Android devices (as many as you want)
-3. Try saying: *"Take screenshot on all devices"*
-4. Watch what happens
+- **Quick Actions panel**: send text, tap an element by name or `x,y`, and launch apps with autocomplete. Arrow keys cycle through what you typed before.
+- **UI Inspector**: hover over any element to see its accessibility ID, resource ID, XPath and bounds.
+- **One-click export** of a selector as a Devicely command, **WebdriverIO**, **Appium Python** or **Appium Java** code.
 
-### **Step 3: Discover the AI layer** *(optional, but fun)*
-- Settings → AI Config → Choose a provider (Gemini/Groq are free)
-- Now you can talk to your devices in plain English
-- *"Launch settings, scroll down, open camera"* - just works
+### Devices that stay connected
 
-**💡 Curious tip:** The more devices you connect, the more interesting it gets.
+- **QR enrollment**: scan a code from the Command Center to pair a device over Wi-Fi — no IP addresses to type.
+- **Health monitoring** spots an unplugged cable, a reboot or a driver failure within seconds and cleans up the session.
+- **Auto-reconnect** brings a device back as soon as it reappears on USB or Wi-Fi. It's on by default; set `DEVICELY_AUTO_RECONNECT=0` to turn it off.
+- **Virtual devices**: start and stop Android emulators and iOS simulators from the UI or the CLI, next to your physical devices.
 
 ---
 
-## 🎯 Usage Examples
+## For AI agents (MCP server)
+
+Devicely includes a [Model Context Protocol](https://modelcontextprotocol.io) server, so coding agents can see and control real devices.
 
-### Natural Language Commands
 ```bash
-# Simple actions
-"take screenshot"
-"launch chrome"
-"go to google.com"
+devicely mcp install claude-code      # prints the `claude mcp add` command to run
+devicely mcp install cursor           # also: vscode, windsurf, gemini, claude-desktop
+devicely mcp doctor                   # checks the backend, token and transports
+```
 
-# Multi-step sequences
-"launch settings scroll up launch camera go to youtube.com press home"
+Or add it to Claude Code by hand:
 
-# Cross-platform (automatically uses correct package IDs)
-"launch settings"  # → com.apple.Preferences (iOS) + com.android.settings (Android)
+```bash
+claude mcp add devicely -- npx -y devicely mcp
 ```
 
-### Voice Commands
-1. Click 🎤 microphone button
-2. Speak: "launch settings scroll up launch camera"
-3. Pause for 2 seconds
-4. Command executes on all selected devices!
-
-### Recording & Replay
-1. Click **Record** button
-2. Perform actions on device
-3. Click **Stop Recording**
-4. Save with name
-5. Replay anytime on any device(s)
-
-### Commander Mode
-1. Connect 2+ devices
-2. First device auto-becomes commander
-3. Click on commander screen to broadcast actions
-4. Use top-left dropdown on mirror to swap commander
-5. All devices stay synchronized automatically
+Two transports are available: **stdio** (`devicely mcp`, which starts the backend if it isn't running) and **Streamable HTTP**, protected by a Bearer token. Run `devicely mcp doctor` to see the HTTP endpoint.
 
----
+### 30 tools
 
-## 🤖 AI Provider Comparison
-
-Devicely supports **7 leading AI providers** with multiple models to choose from:
-
-| Provider | Models | Speed | Free Tier | Best For |
-|----------|--------|-------|-----------|----------|
-| **🤖 OpenAI** | GPT-4 Turbo, GPT-4, GPT-3.5 | 🔥🔥 | ❌ | General purpose, accuracy |
-| **✨ Google Gemini** | 2.0 Flash, 1.5 Pro, 1.5 Flash | 🔥🔥🔥 | ✅ Yes | Cost-effective, multimodal |
-| **🧠 Anthropic Claude** | 3.5 Sonnet, 3 Opus, 3 Haiku | 🔥🔥 | ❌ | Complex reasoning, long context |
-| **🐙 GitHub Copilot** | GPT-4, GPT-3.5 | 🔥🔥 | ❌ | Developers with subscription |
-| **⚡ Groq** | Llama 3 70B, Mixtral, Gemma | 🔥🔥🔥 | ✅ Yes | **Ultra-fast inference** |
-| **🌊 Cohere** | Command R+, Command R | 🔥🔥 | ✅ Yes | Enterprise-grade |
-| **🌬️ Mistral AI** | Large, Medium, Small | 🔥🔥 | ❌ | EU compliance, open-source |
-
-### 💡 Recommendations
-- **Free Users:** Start with Gemini 2.0 Flash or Groq Llama 3
-- **Speed Critical:** Use Groq (fastest inference available)
-- **Best Quality:** Claude 3.5 Sonnet or GPT-4 Turbo
-- **Cost Effective:** Gemini Flash or GPT-3.5 Turbo
-
-Get API Keys:
-- **OpenAI:** https://platform.openai.com/api-keys
-- **Gemini:** https://makersuite.google.com/app/apikey
-- **Claude:** https://console.anthropic.com/account/keys
-- **Groq:** https://console.groq.com/keys
-- **Cohere:** https://dashboard.cohere.com/api-keys
-- **Mistral:** https://console.mistral.ai/api-keys
+| Group | Tools |
+| :--- | :--- |
+| **Observe** | `devicely_snapshot`, `devicely_find_elements`, `devicely_wait_for`, `devicely_get_device_logs`, `devicely_get_action_log` |
+| **Act** | `devicely_tap`, `devicely_long_press`, `devicely_swipe`, `devicely_type_text`, `devicely_press_key`, `devicely_open_url` |
+| **Apps** | `devicely_launch_app`, `devicely_close_app`, `devicely_list_apps`, `devicely_install_app`, `devicely_uninstall_app` |
+| **Devices** | `devicely_list_devices`, `devicely_connect_device`, `devicely_disconnect_device`, `devicely_restart_device`, `devicely_set_passcode`, `devicely_doctor` |
+| **Fleet and flows** | `devicely_commander`, `devicely_run_commands`, `devicely_run_natural_language` |
+| **Virtual devices** | `devicely_launch_virtual_device`, `devicely_stop_virtual_device` |
+| **Recordings** | `devicely_save_recording`, `devicely_list_recordings`, `devicely_replay_recording` |
 
----
+**Prompts** for common jobs: `smoke_test`, `reproduce_bug` and `write_test`.
+**Resources** with live context: `devicely://devices`, `devicely://apps`, `devicely://docs/command-language` and `devicely://docs/agent-guide`.
 
-## 📟 CLI Reference
+### Safety controls
 
-Devicely comes with a powerful command-line interface for quick actions:
+Set these in **Settings → AI Agents** (stored in `~/.devicely/mcp.json`):
 
-| Command | Description |
-| :--- | :--- |
-| `devicely start` | Launch the web-based Command Center. |
-| `devicely list` | List all connected iOS and Android devices. |
-| `devicely doctor` | Verify system requirements (ADB, WDA, etc.). |
-| `devicely exec "<cmd>"` | Execute a natural language command on all devices. |
+- **Destructive tools are off by default**: `install_app`, `uninstall_app`, `restart_device`, `set_passcode` and `stop_virtual_device`. When enabled, they can also require confirmation.
+- **Limit agents to specific devices** (`allowedDevices`), turn off individual tools, and redact sensitive values from output.
+- **Every agent action is logged** to an audit trail.
+
+### REST API
+
+The same capabilities are available over HTTP at `/api/v1`, using the same Bearer token — useful for CI pipelines and custom harnesses. View or rotate the token with `devicely mcp token [--rotate]`; it's saved in `~/.devicely/mcp-token`.
+
+### Example
+
+> *"Take a snapshot of the Pixel, check that the Login button is visible, tap it, and confirm the welcome screen appears."*
+
+The agent calls `devicely_snapshot` and `devicely_find_elements`, taps with `devicely_tap`, and checks the result with `devicely_wait_for`.
 
 ---
 
-## 🆕 Latest Features (v2.2.11 - February 2026)
-
-### 🐛 **Commander Mode Error Fix (v2.2.11)**
-- **Fixed**: "Error starting commander mode" when devices are already connected
-- **Improved**: Commander mode now gracefully handles already-active sessions
-- **Enhanced**: Better error messaging and automatic session management
-- **Smart**: Auto-switches commander between devices without throwing errors
-
-### 👑 Commander Mode with Quick Swap (NEW!)
-- **Real-Time Mirroring:** Live screen mirroring of commander device with follower devices displayed in background
-- **Commander Dropdown:** Top-left dropdown on mirror screen for quick commander device changes
-- **One-Click Swap:** Select any connected device to instantly make it the new commander
-- **Visual Indicators:** Commander badge (👑) on device cards and mirror screen
-- **Auto-Synchronization:** WebSocket-based real-time updates across all components
-- **Touch Broadcasting:** Tap commander screen, all followers execute the same action
-- **Seamless Transitions:** Smooth commander swaps with automatic follower re-assignment
-- **Button Updates:** "Connect All" and "Select All" for clearer bulk operations
-
-### 🤖 Multi-AI Provider Support
-- **7 AI Providers:** OpenAI, Gemini, Claude, Copilot, Groq, Cohere, Mistral AI
-- **Model Selection:** Choose specific models for each provider (GPT-4 Turbo, Claude 3.5 Sonnet, Llama 3 70B, etc.)
-- **Smart Defaults:** Auto-selects best model if not specified (marked with ⭐)
-- **Revamped UI:** Modern Settings interface with provider cards, model dropdowns, and help sections
-- **Free Tiers:** Gemini, Groq, and Cohere offer generous free tiers for testing
-- **Performance Hints:** UI shows speed/cost comparisons (Groq = fastest, Claude = best reasoning)
-
-### ✨ Enhanced AI Capabilities
-- **Multi-Platform Intelligence:** Single AI command generates platform-specific actions for iOS and Android
-- **Simultaneous Execution:** All devices execute at the same time, not sequentially
-- **Error Visibility:** AI errors (quota limits, auth failures) shown inline with results
-- **Response Cleaning:** Automatically removes markdown and explanations from AI output
-- **Cross-Platform Apps:** AI generates both iOS and Android app bundles (Settings, Camera, Chrome, etc.)
-
-### 🎤 Improved Voice Input
-- **Extended Listening:** 2-second silence detection for natural speech
-- **Continuous Speech:** Speak multiple phrases with natural pauses
-- **Live Transcription:** See your words accumulate in real-time
-- **Complex Commands:** "launch settings scroll up launch camera go to google.com press home" works flawlessly
-- **No Duplicates:** Fixed transcript accumulation for clean output
-
-### 🔐 Settings Enhancements
-- **Revamped AI Config Tab:** Professional card-based design with dark mode support
-- **Dynamic API Key Input:** Field changes based on selected provider
-- **Show/Hide Toggles:** Individual toggles for all 7 provider keys
-- **Status Display:** Green success box showing current provider, model, and key preview
-- **Quick Tips Section:** In-UI recommendations for choosing providers
-- **Close Button:** Easy navigation back to main screen
-- **Dark Mode:** Proper color schemes for light and dark themes
-
-### 📝 Recording Improvements
-- **AI Recording Editor:** Edit AI-generated recordings like manual ones
-- **Format Conversion:** Seamlessly handles both string and array command formats
-- **Save & Replay:** Edit, save, and replay AI-generated commands
-- **Status Persistence:** Recording status preserved when navigating between screens
+## CLI reference
+
+| Command | What it does |
+| :--- | :--- |
+| `devicely start` | Start the server and open the Command Center |
+| `devicely start -p <port> --no-browser` | Use a custom port without opening a browser (CI, headless machines) |
+| `devicely start --usb-only` / `--wifi-only` | Only detect devices on USB, or only on Wi-Fi |
+| `devicely start --launch <name>` | Boot a virtual device when the server starts |
+| `devicely start --debug` | Turn on debug logging |
+| `devicely list [--verbose]` | List connected iOS and Android devices |
+| `devicely exec "<command>" [-d <name>]` | Run a command on all devices, or on one with `-d` |
+| `devicely doctor` | Check Node, adb, Xcode and driver setup |
+| `devicely emulators list` | List Android emulators and iOS simulators (alias: `devicely virtual`) |
+| `devicely emulators start <name> [--headless]` | Boot a virtual device |
+| `devicely emulators stop <name>` | Shut down a virtual device |
+| `devicely mcp [--url <url>] [--no-autostart]` | Run the MCP stdio server |
+| `devicely mcp doctor` | Check MCP readiness |
+| `devicely mcp token [--rotate]` | Show or rotate the agent token |
+| `devicely mcp install <client>` | Configure `claude-code`, `claude-desktop`, `cursor`, `vscode`, `windsurf` or `gemini` |
+
+Example commands for `exec`:
+
+```bash
+devicely exec "launch settings"
+devicely exec "launch settings scroll down launch camera"
+devicely exec "open url https://github.com" -d "Pixel 8"
+```
 
 ---
 
-## 📜 License & Privacy
+## Supported platforms
 
-**Proprietary & Confidential.**
-Devicely is currently in a controlled rollout phase. This NPM package provides the binary execution environment. Source code remains private under phase 2 & 3 of our roadmap.
+| | Supported |
+| :--- | :--- |
+| **Host machine** | macOS and Linux, Node.js 18+. Windows isn't supported yet. |
+| **Android** | Phones, tablets and emulators via ADB and UIAutomator2. Wireless debugging needs Android 11+. |
+| **iOS / iPadOS** | iPhones, iPads and simulators via WebDriverAgent. Needs a Mac with Xcode. iOS 16+ also requires Developer Mode. |
 
-**Privacy First:**
-- All API keys stored locally on your machine
-- No telemetry or data collection
-- Commands executed directly on your devices
-- No cloud dependencies for core functionality
+### iOS setup
 
-For enterprise inquiries or commercial access, contact: **devicelyai@gmail.com**
+Physical iOS devices are controlled through WebDriverAgent, which must be signed with your Apple ID or developer team before it can run on a device. Run `devicely doctor` and follow the in-app setup guide; it walks you through installing the iOS tools, enabling Developer Mode and trusting the computer. Android devices and iOS simulators don't need signing.
 
 ---
 
-<div align="center">
+## Privacy and security
 
----
+- **Runs locally.** The Devicely server runs on your machine and talks to your devices directly. Your apps, screenshots and test data aren't uploaded anywhere.
+- **Local AI is an option.** Use Ollama to keep prompts and screen data on your machine. With a cloud AI provider, your commands go to that provider under your own API key.
+- **API keys stay on your machine.**
+- **Agents need a token.** MCP over HTTP and the REST API require a 32-byte Bearer token.
+- **Analytics:** the local server sends no telemetry. The hosted dashboard at [devicely-ai.vercel.app](https://devicely-ai.vercel.app) uses anonymous Vercel page analytics; it connects to your local server and doesn't send device data to Vercel.
 
 ---
 
-## 🤔 **Still curious?**
+## What's new in 2.3.4
 
-Mobile development has always been about juggling multiple devices, hunting for the right selectors, and repeating the same tests over and over. What if there was a more interesting way?
+- **Recordings, rebuilt**: locator-based steps, 10 s auto-wait, coordinate fallback, list and grid views, a details drawer and a step editor.
+- **MCP server and REST API**: 30 tools, prompts and resources, stdio and Streamable HTTP transports, safety policy and an audit log, plus one-command setup for six AI clients.
+- **Faster iOS mirroring** with WebDriverAgent MJPEG streaming and a dedicated WDA port per device.
+- **Device health monitoring and auto-reconnect.**
+- **Virtual devices**: Android emulators and iOS simulators from the UI, CLI and MCP.
+- **Passcode tool**: sets a PIN on Android over ADB and gives guided steps on iOS.
 
-### **Give it a try:**
+---
 
-```bash
-npm install -g devicely
-devicely start
-```
+## License
 
-[📦 NPM Package](https://www.npmjs.com/package/devicely) • [🌐 Try it live](https://devicely-ai.vercel.app) • [📧 Questions: devicelyai@gmail.com](mailto:devicelyai@gmail.com)
+**Free to use. Closed source.** Devicely is free to install and use. The source code is proprietary: you may not copy, modify or redistribute it. See [LICENSE](https://github.com/sureshkumarm8/Devicely/blob/main/LICENSE).
 
-*Built by developers who got tired of the same old mobile testing routine.*
+- **Questions and enterprise inquiries:** [devicelyai@gmail.com](mailto:devicelyai@gmail.com)
+- **Report an issue:** [github.com/sureshkumarm8/Devicely/issues](https://github.com/sureshkumarm8/Devicely/issues)
+
+<div align="center">
 
-**⭐ Star this repo if you find it interesting!**
+**If Devicely saves you time, [star it on GitHub](https://github.com/sureshkumarm8/Devicely).**
 
 </div>