What problem does it solve? Modifying existing images—changing backgrounds, swapping elements, applying new styles, or creating variations—normally requires manual design work. This Skill lets you perform AI-driven image edits programmatically from backend JavaScript/TypeScript code or a CLI, returning base64-encoded results you can save directly. ## Core Features & Use Cases - Text-Guided Image Editing: Send an image URL or base64 data URL plus an edit prompt to the z-ai-web-dev-sdk images.generations.edit endpoint and receive a modified image. - Variations, Style Transfer & Element Replacement: Generate multiple variations of one image, convert photos to painting styles, or replace specific elements while preserving the rest of the composition. - CLI Tool: Run quick edits from the terminal with z-ai image-edit using prompt, image, output, and size flags. - Use Case: Redesign a product photo by replacing its white background with a wooden table scene, then batch-generate three lighting variations for an e-commerce listing. ## Quick Start Use the image-edit skill to change the background of ./photo.jpg to a sunset scene and save the result as ./edited.png.
