What problem does it solve?
It prevents LLM-generated frontends from producing cramped typography, broken grids, and bland or static layouts by enforcing strict design structure and GSAP-driven motion.
Core Features & Use Cases
- AIDA-first page structure: guarantees Navigation (Attention), Hero (Interest), GSAP Scroll interactions (Desire), and a strong Action-oriented footer layout.
- Python-seeded variation: simulates deterministic randomization to vary hero architecture, typography stacks, component architectures, and GSAP paradigms.
- Design-quality constraints: enforces wide 2–3 line H1 flow, bans cheap meta-labels, and requires gapless bento grids using dense auto-flow.
- Advanced GSAP behavior: implements ScrollTrigger pinning/stacking/scrubbing, hover physics, and scroll-driven image/text reveals.
Quick Start
Ask the AI: "Create an AWWWARDS-level landing page using the gpt-taste skill requirements with a cinematic hero, a gapless bento grid, and GSAP ScrollTriggers including pinning and scrubbing text reveals."