Summary

This project benchmarks the automated generation of 100 high-quality landing pages using the Claude Code SDK and Zhipu AI's GLM 4.7 model, powered by a custom ui-ux-pro-max Skill. It targets developers and designers seeking to rapidly prototype diverse website concepts from simple text prompts, demonstrating AI's capability to produce complete, responsive designs with integrated design systems.

How It Works

The core mechanism involves a generate.ts script that processes a list of prompts. It leverages the Claude Code SDK to interact with the GLM 4.7 large language model. GLM 4.7 then utilizes the ui-ux-pro-max Skill, which encapsulates design specifications, component libraries, and best practices, to generate full landing page code, including responsive layouts and animations. This approach aims to translate a single-sentence prompt into a functional website structure.

Quick Start & Requirements

• Installation: npm install
• Prerequisites: Node.js environment, GLM_API_KEY environment variable set with a valid Zhipu AI API key. The generate.ts script may require modification for alternative LLM endpoints or models. Playwright is an optional dependency for browser automation tasks like screenshots.
• Execution: Run the generation process using npm run generate.
• Resources: Generating all 100 pages is noted to be time-consuming; the script supports resuming interrupted runs.
• Links: Zhipu AI Open Platform, Claude Code SDK, GLM 4.7 Compatible API Docs.

Highlighted Details

• Automated generation of 100 landing pages across a wide spectrum of categories including SaaS, E-commerce, Finance, Tech, Healthcare, and Creative industries.
• Employs a custom ui-ux-pro-max Skill, integrating design system rules and component libraries for consistent, high-quality output.
• Demonstrates a "one-sentence prompt to a complete website" workflow, showcasing AI's potential for rapid web prototyping.

Maintenance & Community

The README does not provide specific details on maintainers, community channels (like Discord/Slack), or a public roadmap. Sponsorships or partnerships are not explicitly mentioned, beyond a referral link for purchasing GLM 4.7.

Licensing & Compatibility

• License: MIT License.
• Compatibility: The MIT license generally permits commercial use and integration into closed-source projects, with attribution requirements.

Limitations & Caveats

The project is presented as a benchmark, implying its primary goal is demonstration rather than production-ready deployment. The generation process is computationally intensive and time-consuming. Success is dependent on the quality and availability of the GLM 4.7 API and the effectiveness of the ui-ux-pro-max Skill. Users must manage API keys and potential costs associated with LLM usage.