menuvision

What This Skill Does

MenuVision is a powerful end-to-end automation tool within the OpenClaw ecosystem designed to transform static restaurant information—whether captured from a live URL, an uploaded PDF, or raw photos of a printed menu—into a professional, visually stunning digital experience. By leveraging the advanced multimodal capabilities of Gemini Vision, the skill extracts item names, descriptions, and prices into a standardized structured JSON format. It then utilizes Gemini Image generation to create custom, appetizing food photography tailored to your specific menu items. Finally, the skill assembles a high-performance HTML file with all CSS, JavaScript, and asset references included, ensuring a portable and responsive digital menu.

Installation

To begin using MenuVision, ensure you have the OpenClaw environment properly initialized. Run the following command in your terminal to install the skill package:

clawhub install openclaw/skills/skills/ademczuk/menuvision

Once installed, the agent will automatically recognize the 'menuvision' namespace and trigger the associated extraction, generation, and building scripts when you provide input.

Use Cases

• Small Business Digitization: Restaurants with physical menus can quickly go digital by uploading simple photos of their current offerings.
• Aggregator Content Creation: Food bloggers or reviewers can convert complex PDFs from diverse websites into a unified, clean, and consistent web layout.
• Marketing Automation: Use the AI image generation capabilities to create promotional landing pages for seasonal menus without the need for professional photography.

Example Prompts

1. "Build a menu for https://www.shoyu.at/menus and make sure the layout is mobile-friendly."
2. "Create a photo menu from this PDF [attach file] and include generated images for the signature items."
3. "Make a digital menu from these photos of a local cafe's menu and publish it to my GitHub pages."

Tips & Limitations

• Data Consistency: Always ensure your output matches the defined menu_data.json schema strictly to prevent build errors in the build_menu.py script.
• Quality Control: While Gemini Image is excellent for illustrative purposes, always verify specific allergy information or dietary symbols manually, as OCR may occasionally misinterpret highly stylized fonts in scanned images.
• Processing Time: Generating unique images for every item can be time-consuming; start with a single section to test the visual style before generating an entire menu with 50+ items.