@IndieDevHailey: The aesthetic of this garden-skills is amazing, finally saying goodbye to AI-style websites! ConardLi's garden-skills has 6k+ stars. 25 style recipes directly elevate the aesthetics of AI-generated web pages—no more cookie-cutter blue-purple cards, Linear, Aes…

X AI KOLs Timeline Tools

Summary

ConardLi's garden-skills project has achieved 6k+ stars, offering 25 style recipes and 4 Skill modules. Installable with a single npx command, it supports platforms like Claude Code and Cursor, aiming to enhance the aesthetics of AI-generated web pages and move away from monotonous designs.

The aesthetic of garden-skills is simply superb, finally saying goodbye to AI-style websites! ConardLi's garden-skills has 6k+ stars. With 25 style recipes, it directly takes the aesthetics of AI-generated web pages to the next level—no more monotonous blue-purple cards. Switch seamlessly between Linear, Aesop, Pentagram styles, delivering polished, high-end results. 4 Skill modules, all installable with one npx command, supporting six major platforms including Claude Code, Cursor, and Codex. - web-design-engineer: AI truly understands design systems, color schemes, and typography, directly outputting polished landing pages, dashboards, and prototypes. Built-in anti-template checklist + 25 style recipes - web-video-presentation: One sentence or article → one-click generation of clickable 16:9 web demo videos (23 themes + TTS voiceover) - gpt-image-2: Professional-grade image generation/editing prompt engine, supporting local and multi-platform - kb-retriever: Precise local knowledge base retrieval without blowing up context Go try it out, productivity will take off immediately
Original Article
View Cached Full Text

Cached at: 05/26/26, 05:12 PM

This garden-skills aesthetic is incredible — finally saying goodbye to websites that scream “AI-generated”!

ConardLi’s garden-skills has already hit 6k+ stars.

25 style recipes instantly elevate the look of AI-generated web pages — no more cookie-cutter blue-purple cards. Switch between Linear, Aesop, Pentagram vibes on the fly, and get a polished, premium feel every time.

4 Skill modules, all installable with a single npx command, supporting six major platforms including Claude Code, Cursor, and Codex.

  • web-design-engineer: AI that truly understands design systems, color palettes, and typography — directly outputs polished landing pages, dashboards, and prototypes. Built-in anti-template checklist + 25 style recipes.
  • web-video-presentation: Turn a sentence or article into a clickable 16:9 web presentation video with one click (23 themes + TTS voiceover).
  • gpt-image-2: Professional-grade image generation/editing prompt engine, supporting local and multi-platform use.
  • kb-retriever: Precise retrieval from local knowledge bases without blowing up the context.

Give it a try — your productivity is about to take off.

Similar Articles

@vista8: Open source, open source! Tired of AI-heavy web design and monotonous layouts? Qiao Bangzhu tested and compared 8 mainstream design Skills, absorbed the strengths of each, and open-sourced his own design Skill: qiaomu-design. Features: 1. Style Fitting Room: Generate four design demos from a single sentence, preview and confirm before development.

X AI KOLs Timeline

Qiao Bangzhu open-sourced the design Skill qiaomu-design, offering features like Style Fitting Room, anti-AI-style design, self-evolution mechanism, and references from mature websites, supporting GLM or Claude models.

@xingxingli45573: Another great find called Taste Skill, which has already gained 20.3k stars on GitHub. Simply put, after installing it, the frontend interfaces generated by AI look much better. Layout, fonts, animations, whitespace — all aspects have improved significantly, giving an overall premium feel. It offers several design directions...

X AI KOLs Timeline

Taste Skill is an open-source frontend framework designed to enhance the visual quality of AI-generated interfaces. It offers multiple design directions and adjustable parameters, and has gained 20.3k stars on GitHub.