cloudinary-image

Cloudinary image optimization CLI with Node.js wrappers for select Cloudinary Node.js APIs
covering image transformation, optimization, and asset management.

Requirements

• Node v24+
• Docker (optional)
• Cloudinary account

Installation

Create a .env file in the /app directory, replacing the contents of the .env.example file with actual values.

Variable Name	Description
CLOUDINARY_NAME	Cloudinary account name
CLOUDINARY_API_KEY	Cloudinary API key
CLOUDINARY_API_SECRET	Cloudinary API secret
CHOKIDAR_USEPOLLING	Enables file watching on tsx watch running inside Docker containers on a Windows host. Set it to true if running Docker Desktop with WSL2 on a Windows OS host.
CHOKIDAR_INTERVAL	Chokidar polling interval. Set it along with CHOKIDAR_USEPOLLING=true if running Docker Desktop with WSL2 on a Windows OS host. The default value is 1000.

Usage

A. Using Docker

1. Build the image.
   docker compose build

2. Run the container.
   docker compose up

3. Edit the .ts source files and watch for changes.

4. Run the Available Scripts using Docker.

5. See the examples under the Code Samples section for more information.

Example using the development Docker image

(PowerShell - development)

docker exec cloudinary-cli-dev npm run docker:debug -- -f /opt/app/assets/sunset.jpg -u -d

Example using stand-alone production Docker image

(PowerShell - production)

Build the production image with
docker compose -f docker-compose.prod.yml build

 docker run --rm --env-file .env `
   -v ${pwd}/assets:/images `
   weaponsforge/cloudinary-cli `
   -f /images/sunset.jpg -u

B. Using Node.js

1. Install dependencies.

   cd app
   npm install

2. Run the app in development mode.

   npm run dev
   

3. Edit the .ts source files and watch for changes.

4. Run the Available Scripts.

5. See the examples under the Code Samples section for more information.

Available Scripts

npm start

Optimizes an input image using the Cloudinary image transformations. Downloads the optimized image to a /processed directory relative to the input file, or to a specified output directory.

NOTE: this requires transpiling TypeScript into JavaScript first via npm run build.

Example Usage

npm start -- -f /path/to/file.jpg -u -d

CLI Guide

npm start -- \
  -f /path/to/file.jpg       # Full input image file path
  -o /output/folder/path     # (Optional) output folder
  -a my-asset-folder         # (Optional) Cloudinary asset folder
  -t cars,vehicles,tech      # (Optional) image tags
  -w 600                     # (Optional) width to resize the image. Default is 800
  -u                         # (Optional) flag to upload the input image to Cloudinary. Required on 1st run.
  -d                         # (Optional) flag to delete the uploaded image in Cloudinary

NOTE: This script is also accessible using npx optimize minus the -- flag.

npm run dev

Runs the npm start script in development mode with file watching using tsx.

Example usage:
npm run dev -- -f /assets/sunset.jpg -u

npm run info

Logs the installed Node.js and npm version, environment platform, architecture and V8 version.

npm run build

Builds JavaScript, .d.ts declaration files, and map files from the TypeScript source files in the /src directory to the /dist directory.

npm run types:check

Runs type-checking without generating the JavaScript or declaration files from the TypeScript files in the /src directory.

npm run lint

Lints TypeScript source codes.

npm run lint:fix

Fixes lint errors in TypeScript files.

npm run watch

Watches file changes in .ts files using the tsc --watch option.

npm run docker:dev

Docker counterpart of the npm run dev script. Exports the IS_DOCKER=true variable and runs the npm run dev script in development mode with file watching using tsx within Docker.

Tip

Set CHOKIDAR_USEPOLLING=true and CHOKIDAR_INTERVAL=1000 in the .env file to enable file watching on when running inside Docker containers on a Windows host.

Uncomment and use _values in /src/scripts/optimize/index.ts to manually set optimize(_values) not from CLI input.

npm run docker:watch:win

Watches file changes in .ts files using the tsc --watch option with dynamicPriorityPolling in Docker containers running in Windows WSL2.

🧾 Code Samples

A. Optimize an Image

import { join } from 'node:path'
import dotenv from 'dotenv'
import { CloudinaryImage } from '@/lib/image.js'

dotenv.config()

const main = async () => {
  const filePath = join(process.cwd(), 'boat.jpg')

  const image = new CloudinaryImage({
    localFile: filePath,
    cloudinaryAssetFolder: 'my-folder',
  })

  await image.upload('sea,travel')
  await image.optimize(600)
  await image.delete()
}

main()

B. Apply Image Transformations

import dotenv from 'dotenv'
import { CloudinaryImage } from '@/lib/image.js'
import { join } from 'node:path'

dotenv.config()

const main = async () => {
  const filePath = join(process.cwd(), 'boat.jpg')

  const image = new CloudinaryImage({
    localFile: filePath,
    cloudinaryAssetFolder: 'my-folder',
  })

  // Upload image to Cloudinary
  await image.upload('sea,travel')

  // Generate URL of resized image
  const urlResize = await image.transformer
    .resize(image.publicId, {
      width: 450
    })

  // Generate URL of cropped image
  const urlCropped = await image.transformer
    .crop(image.publicId, {
      width: 400,
      height: 200,
      crop: 'scale'
    })

  // Generate URL of image's new format
  const urlFormat = await image.transformer
    .format(image.publicId, 'webp')

  // Generate URL of image with improved quality
  const urlQuality = await image.transformer
    .quality(image.publicId, 'auto')

  // Download one of the generated images
  const downloadFilePath = join(process.cwd(), image.name)
  await image.service.fetch(urlCropped, downloadFilePath)
}

main()

C. Using Classes

import { join } from 'node:path'

import { AssetManager } from '@/lib/cloudinary/manager.js'
import { AssetService } from '@/lib/cloudinary/service.js'
import { BaseImage } from '@/lib/cloudinary/baseimage.js'
import { Transform } from '@/lib/cloudinary/transform.js'

// Class for managing Cloudinary assets
const _manager = new AssetManager()

// Class for uploading and fetching images from Cloudinary
const _service = new AssetService()

// Class for generating Cloudinary image transformations
const _transformer = new Transform()

// Initialize a new BaseImage - no Cloudinary libraries
const inputFile = join(process.cwd(), 'boat.jpg')
const outputFile = join(process.cwd(), 'images', 'done', 'processed.jpg')

const _image = new BaseImage({
  localFile: inputFile,
  cloudinaryAssetFolder: 'my-folder',
  localDestination: outputFile, // optional
})

// Note: the CloudinaryImage class is composed of all these components

References

• Cloudinary NPM Registry
• Cloudinary Node.js Docs

@weaponsforge
20260817