what is wardrobe fr?

Wardrobe is a JavaScript app that takes photos of your clothes and turns them into an organized, digital closet using OpenAI's image and vision models. You point it at pictures of your outfits, and it detects each individual garment, extracts a clean cutout of that item, and can also generate a modeled preview of how it looks on a reference photo of you. To run it, you clone the repository, install dependencies with npm, copy the example environment file, and start the development server. The importer will not work until you add your own OpenAI API key to the environment file and place a reference photo of yourself, saved as a PNG, in the data folder. Once running, the app is available in a browser at a local address. The project includes two Codex skills for automating the process. One skill imports clothes from a folder of photos, generates modeled item photos, and adds everything to your wardrobe. The other skill takes the clothes already in your wardrobe and generates full outfit ideas, including a modeled lookbook. Both skills ask questions as they go, such as which folder to import from or how many outfit looks to create, and they review each generated image before saving it. Under the hood, garment detection uses the OpenAI Responses API, cutouts are produced with the OpenAI Images API, and everything the app creates, including original photos, background jobs, generated images, and the underlying database file, is kept locally on your own machine rather than sent to a remote server for storage. The app is configured through environment variables, letting you choose which vision and image models to use, the image quality level, and where your model reference photo and data folder live. The app itself supports common actions like dragging and dropping photos in, pasting them, editing entries, reviewing results, regenerating an image you are not happy with, and approving the final version. Wardrobe is released under the MIT license, and requires Node version 22 or newer to run.