본문으로 건너뛰기
cccskills
무료GitHub에 공개

image-optimization-engineer

Optimize web images for performance with Next.js Image, responsive srcset, AVIF/WebP, lazy loading, and blur placeholders. Activate on: image optimization, LCP improvement, responsive images, AVIF, blur placeholder, next/image. NOT for: image generation/editing (use image-gen skills), SVG icon systems (use design-system-creator).

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

설치 방법 보기

포함된 파일(1)

  • SKILL.md7.9 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Image Optimization Engineer

Deliver optimally-sized, modern-format images with lazy loading, blur placeholders, and responsive art direction for fast LCP and minimal bandwidth.

Activation Triggers

Activate on: slow Largest Contentful Paint, large image payloads, missing responsive images, next/image configuration, AVIF/WebP conversion, blur-up placeholders, image CDN setup, srcset and sizes attributes.

NOT for: generating or editing images (use qwen-image, FLUX, or creative tools). SVG icon sprite systems -- use design-system-creator. Video optimization -- different domain entirely.

Quick Start

  1. Audit current images -- run Lighthouse; check LCP element, total image weight, and format usage.
  2. Use next/image -- automatic AVIF/WebP serving, lazy loading, blur placeholders, and responsive srcset.
  3. Set explicit sizes -- tell the browser the rendered width so it picks the right srcset candidate.
  4. Generate blur placeholders -- use plaiceholder or Next.js built-in placeholder="blur" for static imports.
  5. Verify with WebPageTest -- confirm images serve AVIF to supporting browsers and correct sizes at each breakpoint.

Core Capabilities

DomainTechnologiesKey Patterns
Framework Integrationnext/image, Astro <Image>, Vite vite-imagetoolsAutomatic optimization at build/request time
Modern FormatsAVIF, WebP, fallback JPEG/PNGContent negotiation via Accept header
Responsive Imagessrcset, sizes, <picture>Art direction, resolution switching
Lazy Loadingloading="lazy", Intersection ObserverDefer off-screen images
PlaceholdersLQIP (blur), solid color, blurhashPerceived performance during load
CDN/LoadersCloudflare Images, Imgix, CloudinaryOn-the-fly resize, format conversion, caching

Architecture Patterns

Pattern 1: Next.js Image with Proper Sizing

import Image from 'next/image';

// Static import -- enables automatic blur placeholder
import heroImage from '@/public/images/hero.jpg';

export function HeroSection() {
  return (
    <Image
      src={heroImage}
      alt="Product hero shot showing the dashboard in action"
      placeholder="blur"          // automatic blurhash from static import
      priority                     // preload -- this is the LCP element
      sizes="100vw"                // full-width hero
      quality={85}
      className="w-full h-auto object-cover"
    />
  );
}

export function ProductGrid({ products }: { products: Product[] }) {
  return (
    <div className="grid grid-cols-2 md:grid-cols-3 lg:grid-cols-4 gap-4">
      {products.map(product => (
        <Image
          key={product.id}
          src={product.imageUrl}
          alt={product.name}
          width={400}
          height={400}
          sizes="(max-width: 768px) 50vw, (max-width: 1024px) 33vw, 25vw"
          loading="lazy"            // NOT priority -- below the fold
          placeholder="blur"
          blurDataURL={product.blurHash}
        />
      ))}
    </div>
  );
}

Pattern 2: Art Direction with <picture>

When you need different crops/aspect ratios at different breakpoints:

export function ResponsiveHero({ image }: { image: HeroImage }) {
  return (
    <picture>
      {/* Mobile: square crop */}
      <source
        media="(max-width: 639px)"
        srcSet={`${image.mobileUrl}?w=640&format=avif 1x, ${image.mobileUrl}?w=1280&format=avif 2x`}
        type="image/avif"
      />
      <source
        media="(max-width: 639px)"
        srcSet={`${image.mobileUrl}?w=640&format=webp 1x, ${image.mobileUrl}?w=1280&format=webp 2x`}
        type="image/webp"
      />
      {/* Desktop: wide crop */}
      <source
        media="(min-width: 640px)"
        srcSet={`${image.desktopUrl}?w=1200&format=avif 1x, ${image.desktopUrl}?w=2400&format=avif 2x`}
        type="image/avif"
      />
      <source
        media="(min-width: 640px)"
        srcSet={`${image.desktopUrl}?w=1200&format=webp 1x, ${image.desktopUrl}?w=2400&format=webp 2x`}
        type="image/webp"
      />
      {/* Fallback */}
      <img
        src={`${image.desktopUrl}?w=1200&format=jpeg`}
        alt={image.alt}
        loading="eager"
        decoding="async"
        className="w-full h-auto"
      />
    </picture>
  );
}

Image Pipeline Architecture

  ┌─ Source Image (4000x3000 JPEG, 8MB) ──────────────┐
  │                                                     │
  │  Build/CDN Pipeline:                                │
  │  ├─ Resize: 640, 960, 1280, 1920, 2560            │
  │  ├─ Format: AVIF (best), WebP (fallback), JPEG     │
  │  ├─ Quality: 80 (AVIF), 85 (WebP), 85 (JPEG)      │
  │  └─ Generate: blur placeholder (32x32 base64)      │
  │                                                     │
  │  Browser receives (via content negotiation):        │
  │  ├─ Chrome/Edge: AVIF @ matched width (~40KB)       │
  │  ├─ Safari 16+: WebP @ matched width (~60KB)        │
  │  └─ Legacy: JPEG @ matched width (~90KB)            │
  │                                                     │
  │  Result: 8MB → 40-90KB (98% reduction)              │
  └─────────────────────────────────────────────────────┘

Anti-Patterns

  1. Missing sizes attribute -- without it, the browser assumes the image is 100vw and downloads the largest srcset candidate. A 25vw grid image downloads 4x too large.
  2. priority on every image -- preloads them all, defeating the purpose. Only the LCP element (usually the hero image) should have priority.
  3. Serving PNG for photographs -- PNG is lossless and 5-10x larger than AVIF for photos. Use PNG only for graphics with transparency and sharp edges; AVIF/WebP for photos.
  4. No explicit width and height -- causes Cumulative Layout Shift (CLS) because the browser cannot reserve space before the image loads. Always set dimensions or use aspect-ratio.
  5. Blur placeholder on tiny thumbnails -- the blur placeholder adds ~300 bytes of inline base64. For 32x32 thumbnails, that is larger than the image itself. Skip placeholders for small images.

Quality Checklist

  • LCP image has priority (or fetchpriority="high") and is preloaded
  • All images specify sizes attribute matching their actual rendered width at each breakpoint
  • AVIF served to supporting browsers (verify with DevTools Network panel Type column)
  • width and height set on every <img> (CLS score = 0 for images)
  • Off-screen images use loading="lazy" (not priority)
  • Blur placeholders on hero/above-fold images for perceived performance
  • No images over 200KB at their served resolution
  • decoding="async" set on non-critical images
  • Image CDN or next/image handles format negotiation (no manual <picture> unless art direction needed)
  • Lighthouse image audit passes with no "properly size images" or "serve modern formats" warnings

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Expert in 2000s-era music visualization (Milkdrop, AVS, Geiss) and modern WebGL implementations. Specializes in Butterchurn integration, Web Audio API AnalyserNode FFT data, GLSL shaders for audio-reactive visuals, and psychedelic generative art. Activate on "Milkdrop", "music visualization", "WebGL visualizer", "Butterchurn", "audio reactive", "FFT visualization", "spectrum analyzer". NOT for simple bar charts/waveforms (use basic canvas), video editing, or non-audio visuals.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

Expert legal research agent for finding and scraping expungement data state by state. Knows authoritative sources, URL patterns, Firecrawl configuration, and 2026 legal landscape.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

Expert in 3D computer vision labeling tools, workflows, and AI-assisted annotation for LiDAR, point clouds, and sensor fusion. Covers SAM4D/Point-SAM, human-in-the-loop architectures, and vertical-specific training strategies. Activate on '3D labeling', 'point cloud annotation', 'LiDAR labeling', 'SAM 3D', 'SAM4D', 'sensor fusion annotation', '3D bounding box', 'semantic segmentation point cloud'. NOT for 2D image labeling (use clip-aware-embeddings), general ML training (use ml-engineer), video annotation without 3D (use computer-vision-pipeline), or VLM prompt engineering (use prompt-engineer).

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

Apply crisis decision-making research to agent routing, uncertainty triage, and coordination failure analysis in time-pressured systems. Use when diagnosing handoff failures, analytical paralysis, or expert judgment under incomplete information. NOT for routine coding, simple CRUD design, or static single-agent tasks with complete information.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

Use for insight, reframing, contradiction, impasse, and anomaly-driven problem solving when execution effort no longer helps. NOT for routine optimization, error correction, or well-specified tasks with known solution paths.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

Apply normative BDI reasoning to agents that must detect norms, choose which commitments to internalize, and resolve conflicts by comparing consequences. Use when obligations, prohibitions, or policies collide in autonomous systems. NOT for simple fixed-priority rules, pure constraint satisfaction, or domains where no real normative conflict exists.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

curiositech/windags-skills162026년 10월 14일 업데이트

curiositech의 스킬 모두 보기

이 스킬의 문제 신고