Node Native Win Utils

I did it for myself because I didn't feel like dealing with libraries like 'node-ffi' to implement this functionality. Maybe someone will find it useful. It's WINDOWS OS ONLY

Features

• Global keyboard event listener (keyDown / keyUp)
• Window information & screenshots
• Mouse movement, clicks, drag & drop
• Keyboard emulation (keyPress, typeString)
• OpenCV integration (template matching, blur, grayscale, histogram equalization, color manipulation, ROI, drawing)
• Tesseract OCR text recognition
• Screen capture
• Mouse event listener
• Prebuilt binaries (no Python or build tools required on Windows)
• ESM + CommonJS support

Requirements

• Windows 10 or later (x64)
• Node.js >= 18 (prebuilts for recent versions via prebuildify)

Installation

npm install node-native-win-utils

Usage

import {
  KeyboardListener,
  KeyCodeHelper,
  KeyListener,
  getWindowData,
  captureWindow,
  captureWindowN,
  captureScreenToFile,
  mouseMove,
  mouseClick,
  mouseDrag,
  typeString,
  keyPress,
  textRecognition,
  OpenCV,
  startMouseListener,
  stopMouseListener,
} from "node-native-win-utils";

Keyboard

// Singleton listener
const listener = KeyboardListener.listener();
listener.on("keyDown", (data) => console.log("Down:", data.keyName));
listener.on("keyUp", (data) => console.log("Up:", data.keyName));

// Simulate key press
keyPress(KeyCodeHelper.A);           // once
keyPress(KeyCodeHelper.Enter, 2);    // twice

Mouse & Typing

mouseMove(500, 300);
mouseClick();                    // left click
mouseClick("right");
mouseClick("left", "down") // left button down
mouseClick("left", "up") // left button up
mouseDrag(100, 100, 800, 600, 50); // optional speed

typeString("Hello from Node!", 30); // 30ms delay per char

Window Operations

const data = getWindowData("Notepad");
console.log(data); // { width, height, x, y }

captureWindow("Notepad", "screenshot.png");
const buffer = captureWindowN("Notepad"); // Buffer

Screen Capture

await captureScreenToFile("full-screen.png");

Text Recognition (Tesseract OCR)

import path from "path";

const text = textRecognition(
  path.join(__dirname, "traineddata"), // path to .traineddata files
  "eng",
  path.join(__dirname, "image.png")
);
console.log(text);

OpenCV (image processing)

import { OpenCV } from "node-native-win-utils";

const img = new OpenCV("image.png"); // or ImageData { width, height, data: Uint8Array }

// Chainable methods
const processed = img
  .blur(5, 5)
  .bgrToGray()
  .equalizeHist()
  .darkenColor([240, 240, 240], [255, 255, 255], 0.5) // lower, upper, factor
  .drawRectangle([10, 10], [200, 100], [255, 0, 0], 3)
  .getRegion([50, 50, 300, 200]);

processed.imwrite("processed.png");

// Template matching
const match = img.matchTemplate(templateImage, /* method */, /* mask */);
console.log(match); // { minValue, maxValue, minLocation, maxLocation }

Mouse Event Listener

startMouseListener(({ x, y, type }) => {
  console.log(`${type} at ${x},${y}`); // move at 400 300
});

mouseClick("left"); // trigger event

// terminates a thread
stopMouseListener(); 

Building from source

npm install
npm run build

Requires:

Visual Studio Build Tools (C++).
Python 3

License

--

OpenCV License

MIT License

Made with ❤️ for Windows automation.
Issues / PRs welcome!