responsive-images

skill:responsive-images - Performant Responsive Images

Version: 1.0.0

Purpose

Implement performant responsive images that load the right size and format for every device. This skill covers srcset /sizes attributes, modern formats (WebP, AVIF), lazy loading, CLS prevention with aspect-ratio , the <picture> element for art direction, fetchpriority for LCP images, and framework-specific image components (Next.js Image , Astro Image ). Use when adding images to any web project, optimizing existing images for performance, or fixing Core Web Vitals (CLS, LCP) issues related to images.

File Structure

skills/responsive-images/ ├── SKILL.md (this file) └── examples.md

Interface References

Context: Loaded via ContextProvider Interface
Memory: Accessed via MemoryStore Interface
Shared Patterns: Shared Loading Patterns
Schemas: Validated against context_metadata.schema.json and memory_entry.schema.json

Image Optimization Focus Areas

Resolution Switching: srcset with width descriptors (w ) and sizes attribute for viewport-based selection
Format Selection: <picture> with <source> for AVIF → WebP → JPEG/PNG fallback chain
Art Direction: <picture> with media queries for different crops at different breakpoints
Lazy Loading: loading="lazy" for below-fold images, loading="eager" for LCP images
Fetch Priority: fetchpriority="high" on LCP image, fetchpriority="low" on non-critical
CLS Prevention: Explicit width /height or aspect-ratio CSS to reserve space before load
Decoding: decoding="async" for non-critical images to avoid blocking main thread
Framework Components: next/image , Astro <Image> , Nuxt <NuxtImg> — automatic optimization

Common Pitfalls Prevented

Missing width and height attributes causing CLS (Cumulative Layout Shift)
Using loading="lazy" on the LCP (Largest Contentful Paint) image — slows it down
srcset without sizes — browser defaults to 100vw , downloading oversized images
sizes="100vw" on images that are never full-width (sidebar images, cards)
Missing AVIF/WebP sources — serving large JPEG/PNG to modern browsers
fetchpriority="high" on too many images — defeats the purpose
Lazy loading images that are in the initial viewport
Missing alt text on informational images (accessibility violation)
Using CSS background-image for content images (not accessible, not optimizable)
Serving retina images to non-retina screens (wasted bandwidth)

MANDATORY WORKFLOW (MUST FOLLOW EXACTLY)

Step 1: Initial Analysis

YOU MUST:

Identify the image context:
Hero/banner images (LCP candidates)
Content images (articles, products)
Thumbnails/avatars
Decorative/background images
Detect the framework and any image optimization tooling
Identify the image source (local files, CMS, CDN, user uploads)
Determine target breakpoints from the project's responsive design
Ask clarifying questions:
What image formats are available?
Is there an image CDN (Cloudinary, imgix, Vercel Image Optimization)?
What are the Core Web Vitals targets?

Step 2: Load Memory

Follow Standard Memory Loading with skill="responsive-images" and domain="engineering" .

YOU MUST:

Use memoryStore.getSkillMemory("responsive-images", "{project-name}") to load project patterns
Use memoryStore.getByProject("{project-name}") for cross-skill context (breakpoints, design system)
Review previously documented image patterns, CDN config, and format support

Step 3: Load Context

Follow Standard Context Loading for the engineering domain. Stay within the file budget declared in frontmatter.

Step 4: Implement Responsive Images

YOU MUST apply these patterns:

LCP Hero Image (above-fold, critical):

Content Image (below-fold):

Modern Format Chain:

CLS Prevention:

.image-container { aspect-ratio: 16 / 9; width: 100%; overflow: hidden; } .image-container img { width: 100%; height: 100%; object-fit: cover; }

Rules:

LCP image: fetchpriority="high" , loading="eager" (or omit loading), no lazy loading
Below-fold images: loading="lazy" , decoding="async"
Always provide width and height OR CSS aspect-ratio
sizes must reflect the actual rendered width at each breakpoint
Every <img> with content meaning must have descriptive alt text
Decorative images: alt="" and role="presentation"

DO NOT use loading="lazy" on LCP images

Step 5: Generate Output

Save to /claudedocs/responsive-images_{project}_{YYYY-MM-DD}.md
Follow naming conventions in ../OUTPUT_CONVENTIONS.md
Include HTML/JSX code, CSS, and any configuration changes

Step 6: Update Memory

Follow Standard Memory Update for skill="responsive-images" .

Store image conventions, breakpoints used, CDN configuration, format support, and component patterns.

Compliance Checklist

Before completing, verify:

Step 1: Image context and optimization scope identified
Step 2: Standard Memory Loading pattern followed
Step 3: Standard Context Loading pattern followed
Step 4: Responsive images implemented with correct srcset/sizes, lazy loading, and CLS prevention
Step 5: Output saved with standard naming convention
Step 6: Standard Memory Update pattern followed

responsive-images

Safety Notice

Copy this and send it to your AI assistant to learn

Source Transparency

Related Skills

office

jquery-4

accessibility