Restaurant Menu Design Generation

Problem

Restaurant pages, delivery listings, and in-store signage often need a clean menu image that combines the restaurant name, dish list, prices, brand colors, and layout ratio. Manual design work can be repetitive: checking Chinese typography, price alignment, visual hierarchy, and color harmony, especially when a brand logo already defines the identity. This skill treats that workflow as a structured generation pipeline.

How it works

The main inputs are restaurant_name, menu, style, and ratio, with an optional logo_url. The skill first performs compliance checks. If a logo is provided, it analyzes the image to extract the dominant color, RGB value, and text contrast suggestions. It then builds a simplified-Chinese layout prompt from the dish text, restaurant name, style, and palette, and calls an image model to render the menu. Real runs are asynchronous: execute_task.py submits the job and returns only a task_id, while query_task.py polls until the final image URL and prompt are available. Without an app key, a one-time simulated response is available for preview only.

Boundaries

Real tasks are strictly asynchronous and do not support synchronous mode. menu and restaurant_name are required, and the original menu text should be preserved to avoid accidental rewriting before generation. If no logo is supplied, the skill relies on the provided style or defaults. Because output depends on an image model, final Chinese typography, price clarity, layout details, and color accuracy still need manual review.