GChenSi-2/ai-metadata-viewer ? reverse-engineered prompt

Build me a small web app where I can drag and drop an AI generated image and instantly see the hidden metadata in a clean, easy to read way.

It should work fully in the browser with no upload and no account, and it should read PNG and JPEG files made by Stable Diffusion, ComfyUI, NovelAI, and Midjourney. I want it to pull out useful things like the prompt, model, sampler, seed, LoRAs, and any ComfyUI workflow data, instead of dumping raw text everywhere.

Please also add a simple local history so I can revisit the last few files I checked, without storing the images themselves. It would be nice if the app supports English, Chinese, and Japanese, and feels polished enough for a public site with good SEO and shareable pages.

Use a modern React and TypeScript setup, keep everything client side, and make sure it works well on Vercel. If you need to look up current docs online, go ahead.