What problem does it solve? Removing objects, watermarks, or blemishes and replacing specific regions of an image requires precise local edits that generic text-to-image tools cannot target. This Skill routes mask-based edits to a dedicated inpainting endpoint and falls back to description-based edit models when no mask exists. ## Core Features & Use Cases - Mask-Driven Inpainting: Sends a source image plus a grayscale mask (white = inpaint, black = preserve) to Z-Image Turbo Inpainting with tunable strength and control-scale parameters. - Description-Based Fallback: Uses Nano Banana 2 Edit, GPT Image 2 Edit, or FLUX Kontext Pro when the target region must be described in prose instead of a mask. - LoRA Style Support: Applies fine-tuned brand-style LoRA adapters during inpainting via the Z-Image Inpainting LoRA variant. - Use Case: Remove overhead cables from a street photo by supplying the photo URL, a mask of the cables, and a prompt like "Remove overhead cables; preserve rooflines and sky gradient" with strength 0.5. ## Quick Start Ask the AI to remove a specific object or watermark from your image by providing the image URL and a mask URL, and it will run the inpainting job through the runcomfy CLI.
