$ show Leonxlnx/taste-skill
Taste-skill is not a software application in the traditional sense; it is a portable library of AI-agent 'skills' - extremely detailed Markdown instruction files (SKILL.md) distributed via the `npx skills add` CLI (Vercel agent-skills spec) and a .claude-plugin marketplace. The repo's product is the prompt content itself: 13 frontend design-direction skills that agents like Claude Code, Codex, and ChatGPT load to avoid generic 'AI slop' output. The flagship is `taste-skill` v2 (`design-taste-frontend`), a ~5000-line instruction document that codifies anti-pattern bans (zero em-dashes, no section-number eyebrows, no three-equal-card rows), a brief-inference protocol, design-system mapping to official packages (Fluent, Carbon, Polaris, GOV.UK), canonical GSAP/scroll-reactive code skeletons, and a hard pre-flight checklist with mechanical pass/fail criteria. It is complemented by image-generation-only skills (brandkit, imagegen-frontend-web/mobile) that produce reference comps for downstream coding agents, plus variant skills (brutalist, minimalist, soft, stitch, output-enforcement). The repository also contains a `research/laziness/` folder documenting LLM output-truncation causes and remediation, and a small set of Node/Sharp image-processing scripts used to build the README assets. Those scripts contain one genuinely valuable, extractable algorithm: a flood-fill outer-background removal with auto-cropping to content bounds. The scripts are one-off (hardcoded `C:/Users/User/Downloads` source paths, tightly coupled to specific sponsor assets), so they are not drop-in reusable without adaptation, but the background-removal technique itself is portable. This is a content/instruction repository rather than a codebase; software-engineering quality bars (tests, CI) do not apply. Its craftsmanship lives in the rigor, specificity, and mechanical enforcement of the prompt files, which is exceptional for the prompt-engineering domain. The active development cadence, current dependency references (Tailwind v4, Motion/react, React Server Components), and 92k-star traction confirm it is the leading project in its niche.
The LLM identified 4 files worth highlighting for reuse:
The flagship v2 skill: a self-contained instruction document an AI agent loads to produce premium, non-templated frontends. Encodes a brief-inference protocol, three tuneable dials (DESIGN_VARIANCE / MOTION_INTENSITY / VISUAL_DENSITY) with inference matrices, a brief-to-official-package design-system map, canonical GSAP ScrollTrigger code skeletons, and a 60-point pre-flight checklist with mechanical bans (zero em-dashes, eyebrow caps, no false serif/premium-consumer-palette defaults). Reusable as a template for anyone authoring high-quality agent instruction files; demonstrates non-obvious skill in encoding qualitative taste as verifiable rules.
A domain audit protocol for upgrading existing websites/apps. Provides a structured, category-by-category audit (typography, color, layout, states, content, components, code quality) plus an upgrade-technique catalog (variable-font animation, glassmorphism internals, scroll-driven reveals) and a risk-ordered fix-priority list. Isolated and self-contained - it encodes no project-specific logic and could be applied to any frontend codebase. Demonstrates strong skills in turning subjective design quality into concrete, detectable failure patterns.
A Node/Sharp utility that back-removal-transparentizes PNG buttons by flood-filling from the image border (thresholding near-black/near-transparent pixels), then auto-crops to the content bounding box with padding, and re-encodes to WebP. The flood-fill + getBounds + crop pipeline is a genuinely reusable image-cleaning algorithm that could be extracted into any project needing cheap chroma/background removal from UI assets. Non-obvious skill: efficient iterative queue-based flood fill over raw RGBA buffers rather than naive pixel scanning, plus correct alpha-aware bounds computation. The hardcoded C:/Users/User/Downloads source paths limit portability, but the core algorithm transfers cleanly.
Defines a mandatory image-first workflow for visual website tasks: generate reference images first, deeply analyze them, then implement frontend to match. Adds anti-drift discipline (preserve layout/spacing/typography from the reference rather than re-interpreting generically) and section-first generation rules (one image per section, never crop old images, regenerate for clarity). Encodes a non-obvious, reusable pipeline for any image-to-code agent integration: generation -> structured extraction -> faithful implementation.