nomo-e2e 0.9.2

Agent tool for controlling desktop app WebViews via Chrome DevTools Protocol (CDP) and WebDriver

NomoE2E

Control WebViews in desktop applications through a unified API, supporting both CDP and WebDriver protocols

🎯 Project Goals

Provide AI Agents with a practical, scalable, and cross-platform consistent desktop WebView automation capability.

• 🔗 Dual Protocol Support: Automatic switching between CDP and WebDriver.
• 🖱️ Unified API: click, type_text, screenshot, execute_script, etc.
• ✅ Out of the Box: Built on Playwright, no need to worry about protocol differences.
• 📸 Observability: Error collection, diagnostic snapshots, screenshots.
• 🚀 Global Installation: Like npm, install with one command and use anywhere.

📦 Supported Scenarios

Environment	Protocol	Description
Electron	CDP	Use --remote-debugging-port
Windows WebView2	CDP	Enable remote debugging
Tauri (tauri-plugin-webdriver)	WebDriver	Standard WebDriver endpoint

🚀 Quick Start

Installation (Recommended)

# Install globally with uv (like npm i -g)
uv tool install -e .
# or Install from PyPI
uv tool install nomo_e2e

# After installation, use directly from any directory
nomo-e2e --help

Or use pip

# Install from source
pip install -e .

# Or from PyPI (after release)
pip install nomo-e2e

Install Playwright Browser Driver

# Required for CDP/WebDriver connection
playwright install chromium

CLI Quick Start

# CDP mode (Electron/WebView2) - Auto-detects CDP endpoints
nomo-e2e open http://localhost:9222
nomo-e2e snapshot
nomo-e2e click "#login"
nomo-e2e fill "#username" "admin"
nomo-e2e screenshot --filename result.png

# WebDriver mode (Tauri2) - Requires --driver flag
nomo-e2e open --driver webdriver http://localhost:4445
nomo-e2e click "button.submit"
nomo-e2e eval "document.title"

# View all commands
nomo-e2e --help

For the complete CLI command reference, see the Skills Package.

Python API

from nomo_e2e import WebViewController

# CDP mode (Electron/WebView2) - Auto-detect
controller = WebViewController.connect("http://localhost:9222")

# Or specify driver explicitly
controller = WebViewController.connect(
    "http://localhost:9222",
    driver="cdp"
)

# WebDriver mode (Tauri)
controller = WebViewController.connect(
    "http://127.0.0.1:4445",
    driver="webdriver"
)

# Perform actions
controller.click("#submit-btn")
controller.type_text("#username", "admin")
controller.screenshot("/tmp/page.png")

# Get page info
print(controller.title)
print(controller.url)

# Diagnostic snapshot
snapshot = controller.diagnostics_snapshot()

# Disconnect
controller.disconnect()

Using Context Manager

from nomo_e2e import WebViewController

# Automatic connection lifecycle management
with WebViewController.connect("http://localhost:9222") as controller:
    controller.click("#login")
    controller.type_text("#username", "admin")
    controller.screenshot("./login.png")
# Automatically disconnected

🔧 Environment Configuration

Enable CDP for Electron

Electron apps only need to add the --remote-debugging-port parameter at startup to enable the CDP debugging port.

Startup command:

# macOS/Linux
npx electron . --remote-debugging-port=9222

# Windows
.\node_modules\.bin\electron . --remote-debugging-port=9222

Configure in package.json:

{
  "scripts": {
    "start:dev": "electron . --remote-debugging-port=9222"
  }
}

Verify CDP port:

# Check if port is open
lsof -i :9222

# Or access the debug endpoint
curl http://localhost:9222/json

E2E test connection:

# Auto-detect CDP
nomo-e2e open http://localhost:9222

# Or specify explicitly
nomo-e2e open --driver cdp http://localhost:9222

---

Enable WebDriver for Tauri2

Tauri2 requires installing the tauri-plugin-webdriver plugin and enabling the WebDriver endpoint.

1. Install the Plugin

# Add plugin (interactive)
cd your-tauri-app
cargo tauri add webdriver

# Or manually add dependency to Cargo.toml
[dependencies]
tauri-plugin-webdriver = "2"

2. Configure the Plugin

src-tauri/src/lib.rs:

use tauri_plugin_webdriver::init;

fn run() {
    tauri::Builder::default()
        .plugin(init())
        .run(tauri::generate_context!())
        .expect("error while running tauri application");
}

3. Configure Permissions

src-tauri/capabilities/main.json:

{
  "permissions": [
    "core:default",
    "webdriver:default"
  ]
}

4. Start Dev Server

Tauri2 dev mode automatically enables the WebDriver endpoint:

# Start dev server (default port 4445)
cd your-tauri-app
cargo tauri dev

# Or specify port
WEBDRIVER_PORT=4445 cargo tauri dev

Note: WebDriver port is only enabled in debug/dev builds, release builds will not expose it.

5. Verify WebDriver Endpoint

# Check if endpoint is ready
curl http://127.0.0.1:4445/status

# Expected response
# {"value":{"ready":true,"message":"Ready."}}

6. E2E Test Connection

nomo-e2e open --driver webdriver http://localhost:4445

---

Quick Reference Table

Application	Protocol	Startup Parameter	Endpoint URL	Auto-detect
Electron	CDP	--remote-debugging-port=9222	http://localhost:9222	✅
Tauri2	WebDriver	cargo tauri dev (auto)	http://127.0.0.1:4445	❌
WebView2	CDP	--remote-debugging-port=9222	http://localhost:9222	✅

📁 Project Structure

nomo-e2e/
├── src/
│   └── nomo_e2e/
│       ├── __init__.py          # Public exports
│       ├── controller.py        # Unified controller
│       ├── adapters/            # Protocol adapters
│       │   ├── base.py         # Abstract base class
│       │   ├── cdp.py          # CDP adapter
│       │   └── webdriver.py    # WebDriver adapter
│       ├── cli/                 # CLI commands
│       ├── exceptions.py        # Exception definitions
│       ├── models.py            # Data models
│       └── skills/              # Skills package
│           └── nomo_e2e_skills/
│               ├── SKILL.md     # Main skill package
│               └── references/   # Reference docs
├── tests/                      # Tests
├── INSTALL.md                  # Installation guide
├── ROADMAP.md                 # Development roadmap
└── AGENTS.md                  # AI development guidelines

📖 Documentation

Document	Description
INSTALL.md	Detailed installation and usage guide
src/nomo_e2e_skills/	Skills package (for LLM Agents)
ROADMAP.md	Development roadmap and milestones
AGENTS.md	AI development guidelines

🔧 Development

# Install dev dependencies
uv sync

# Code linting
uv run ruff check src/ tests/

# Code formatting
uv run ruff format src/ tests/

# Run tests
uv run pytest tests/ -v

# Type checking
uv run mypy src/

⚠️ Notes

• WebDriver port is only allowed in debug/dev builds
• Electron's --remote-debugging-port can be used in any build
• CDP endpoints (like localhost:9222) are auto-detected as CDP protocol
• WebDriver endpoints require --driver webdriver flag
• Ensure you have proper authorization before using on any application
• Some applications may have anti-debugging mechanisms

📝 License

Apache License 2.0 - see LICENSE for details.

Copyright (c) 2026 NomoAiT