---

name: nepa-image-workspace description: AI-powered image editing workspace for the NEPA AI platform. Handles upscaling, background removal, style transfer, image generation, color adjustment, and batch processing. Use when working with images, photo editing tasks, or AI image generation in NEPA AI.

Image Workspace

2,433 lines — 46 methods. The most complete open-source AI image toolkit.

Prerequisites

pip install pillow opencv-python numpy torch torchvision openai rembg realesrgan basicsr psd-tools requests scipy

C++ acceleration: libml.a (NEPA Omni ML) — included in the AXON product ZIP.

Source

/home/billk/projects/nepa-ai-monorepo.BAK/vscode_forked/nepa-ai-backend/image_workspace.py

Setup

cd ~/projects/nepa-ai-monorepo.BAK/vscode_forked/nepa-ai-backend
python run_nepa.py

Backend runs at http://localhost:8765.

Key Methods (46 total)

| Method | Description | |--------|-------------| | upload_image(path, project) | Upload image to project | | upscale_image(path, scale) | Real-ESRGAN 2x/4x/8x upscaling | | remove_background(path) | Background removal (rembg) | | restore_face(path) | GFPGAN/CodeFormer face restoration | | generate_image(prompt) | DALL-E 3 / SD generation | | style_transfer(path, style) | Neural style transfer | | color_grade(path, preset) | Color grading pipeline | | denoise_image(path) | AI denoising | | deblur_image(path) | Deblur/sharpen | | batch_process(paths, ops) | Batch operation pipeline | | export_psd(path) | Export as PSD with layers | | get_dominant_colors(path) | Extract color palette | | strip_exif(path) | EXIF privacy strip | | resize_image(path, w, h) | Smart resize/crop | | composite_images(base, overlay) | OpenCV compositing |

Example Usage

import asyncio
from image_workspace import ImageWorkspace

async def main():
    ws = ImageWorkspace()
    
    # Upscale a BMX screenshot 4x
    result = await ws.upscale_image("bmx_frame.jpg", scale=4)
    
    # Remove background for thumbnail
    result = await ws.remove_background("rider.jpg")
    
    # Generate promo art
    result = await ws.generate_image("BMX rider backflip, sunset, cinematic, 4K")
    print(result["url"])

asyncio.run(main())

API Endpoints

POST /api/image/upload
POST /api/image/upscale
POST /api/image/remove-background
POST /api/image/restore-face
POST /api/image/generate
POST /api/image/style-transfer
POST /api/image/batch
POST /api/image/export-psd

AXON Store

Available at: https://axon.nepa-ai.com (image-workspace, $97) Download: /downloads/image-workspace.zip

---

Full version available at https://axon.nepa-ai.com/products

This skill is part of the NEPA AI Axon workspace suite. Buy once, use forever. No subscription required.