This is a straightforward wrapper around OpenRouter's image generation APIs, specifically FLUX.2 and Gemini 3.1 Flash Image Preview. You call a Python script with a text prompt and it spits out a PNG. It handles both generation and editing if you pass an input image. The documentation is clear about when to use this versus their scientific-schematics skill, which is helpful since that's a real decision point. Setup requires an OpenRouter API key in a .env file. It's basic glue code, but sometimes that's exactly what you need when you want to generate images from the command line without writing your own base64 encoding logic.
npx skills add https://github.com/k-dense-ai/scientific-agent-skills --skill generate-image