Sometimes a page needs an image you don't have. A diagram, an illustration, a placeholder for a design. Instead of switching to another tool, you can now describe what you want in yoinko's chat drawer and it'll generate it for you.
The feature works with DALL-E (via OpenAI), Imagen (via Google), or any image generation endpoint you configure. The generated image is automatically saved as an asset attached to the current page.
How it works: open the chat drawer, describe the image you want (e.g. 'a minimal line drawing of a folder tree'), and yoinko sends the prompt to your configured image provider. The result appears in the chat, and you can insert it into the page with one click.
Under the hood, the server handles the API call and stores the resulting image in data/uploads/. The image is referenced in your page content as a standard asset — no lock-in, no special format.
This is still early — we're working on in-editor image generation (describe → generate → insert inline) for a future release. But for now, the chat-based flow covers most use cases nicely.