scientific-schematics

SUSPICIOUS. The skill’s core behavior broadly matches its stated purpose of diagram generation, and there is no downloader, hidden payload, or obvious credential theft logic. However, it routes model access through OpenRouter rather than official Google endpoints, forwards prompts and API credentials to that intermediary, and uses model branding that does not cleanly match official naming. This is better classified as a medium-risk third-party proxy/data-flow concern than confirmed malicious behavior.