Omage - AI Image Inpainting Assistant

Capture and inpaint images in AI chat streams

Omage transforms your AI chat experience by enabling seamless image editing and inpainting directly within your browser. Designed for users who work with multimodal AI models (like **Google Gemini** and **ChatGPT**), Omage eliminates the tedious process of downloading, editing, and re-uploading images just to make small corrections. ### Key Features * **📸 Instant Capture**: Automatically detects images in your AI chat stream. Simply hover over any image to start editing immediately. * **🎨 In-Place Editor**: A focused overlay editor allows you to crop, mask, and select specific regions of interest without ever leaving the page. * **🤖 Smart Prompt Injection**: Automatically constructs prompts with precise coordinate data and your edit instructions, ensuring the AI understands exactly *where* and *what* to change. * **💡 Dual Selection Modes**: Switch between **Box Boundary** (coordinates) and **Color Mask** (painting) modes. *Tip: AI models may respond differently to each method, so try switching modes if you don't get the desired result!* * **⚡ Streamlined Workflow**: * **Auto-Copy**: The selected/cropped image area is automatically copied to your clipboard. * **Auto-Input**: Your textual prompt is automatically injected into the chat input box. * **📂 Recent History Dock**: A convenient side widget keeps track of your recently captured images, allowing for quick access and iterative editing. * **🔒 Local & Secure**: Operates directly on the page content. All processing logic is executed locally within your browser. ### Privacy & Permissions Omage respects your privacy. Here is a breakdown of the permissions we request and why: * **`storage`**: Used to save your extension preferences (e.g., enabled/disabled state). * **`clipboardWrite`**: Essential for copying your processed images directly to the clipboard so you can paste them into the chat. * **`activeTab`**: Allows the extension to detect images and inject the editor interface only on the specific tab you are currently using. * **`scripting`**: Required to inject the content script (the editor overlay) into the webpage when you activate the extension. * **`declarativeNetRequest` & `host_permissions`**: Used strictly to bypass anti-hotlinking protections (Referer checks) on certain image servers. This ensures that the images you select can be successfully loaded into the editor canvas for modification. Omage does **not** collect or transmit your browsing data. ### How to Use 1. **Activate**: Click the extension icon on the page where you want to use Omage. 2. **Select**: Hover over any image in the chat stream and click the **"Omage"** button, or upload an image via the side dock. 3. **Target**: In the editor, drag the selection box to highlight the area you wish to modify. 4. **Instruct**: Type your modification request (e.g., "Change the red car to blue"). 5. **Process**: Click **"Generate & Copy"**. * The processed image is copied to your clipboard. * The context-aware prompt is inserted into the chat bar. 6. **Finish**: Simply paste (Ctrl+V / Cmd+V) the image into the chat and hit Send! ### Perfect For * AI Artists refining generated details. * Developers testing multimodal capabilities. * Anyone looking to fix artifacts in AI images quickly and efficiently.