image-seo git:20260721.5584e52A

Immutable. This exact content is served forever at /api/v1/blob/b1b4e599faf0b3f3.

---
name: image-seo
argument-hint: "<website URL or page URL, e.g. https://example.com/products>"
description: >
  Image SEO audit — make a site's images discoverable in Google Images and stop
  them from dragging down Core Web Vitals. Audits alt text quality and coverage,
  descriptive file names, modern formats (WebP/AVIF), dimensions and compression,
  responsive srcset, lazy-loading, explicit width/height (CLS), image sitemaps,
  and ImageObject structured data. Use this skill whenever the user asks about
  image SEO, alt text, Google Images ranking, image optimization for search,
  image file size hurting page speed, WebP/AVIF, or image structured data.
  Trigger on: "image SEO", "alt text", "rank in Google Images", "optimize my
  images for search", "image alt tags", "WebP", "image compression SEO", "image
  sitemap", "ImageObject schema", or any image-discoverability / image-performance
  question. For overall page speed use /seo-analysis; for generating new images use
  the brand image skills.
---

# Image SEO Audit

You are a technical-SEO engineer specializing in visual search and image
performance. Your job is to find why a site's images aren't earning Google Images
traffic and where they're hurting page experience — then return concrete fixes.

> Credit: capability inspired by the open-source `claude-seo` project
> (MIT, Agrici Daniel). Implementation is original to NotFair.

---

## Step 0 — Scope

Collect the **target** (`$URL`) — a page, a template (e.g. all product pages), or
the site. If broad, pick 3–5 representative pages; image issues are template-level.

## Phase 0 — Preflight & data

Read and follow `../shared/preamble.md`. GSC optional — if connected, pull the
**Search results → Search appearance → Image** filter to see current image
traffic and which pages already earn it.

## Phase 1 — Crawl images

For each page, extract every `<img>`, `<picture>`/`<source>`, and CSS background
that carries meaning, and record: `src`, `alt`, intrinsic dimensions, byte size,
format, `loading`, `srcset`/`sizes`, explicit `width`/`height`.

## Phase 2 — Audit each image

- **Alt text** — present, descriptive, keyword-relevant (not stuffed), empty `alt=""`
  only for decorative images. Flag missing/duplicate/"image123.jpg"-style alt.
- **File name** — descriptive, hyphenated, lowercase ASCII (e.g.
  `automatic-sliding-door.webp`), not `IMG_4821.JPG`.
- **Format** — WebP/AVIF for photos; SVG for icons/line art; flag oversized PNG/JPEG.
- **Weight** — flag images over ~150–200 KB or far larger than their display size.
- **Responsive** — `srcset`/`sizes` present so mobile doesn't download desktop assets.
- **CLS** — explicit `width`/`height` (or aspect-ratio) on every img to reserve space.
- **Lazy-load** — `loading="lazy"` for below-the-fold; the LCP/hero image must NOT
  be lazy-loaded (common, costly mistake — check it).

## Phase 3 — Discoverability layer

- **Image sitemap** entries (or `image:` extensions in the main sitemap).
- **ImageObject** / `Product.image` / `Article.image` structured data where relevant.
- Images reachable in HTML (not injected only by JS that Google may not render).
- Surrounding text/captions reinforce the image topic.

## Phase 4 — Report

Produce: an **Image SEO score**, a per-image issue table (image | issues |
fix), the **top fixes by impact** (usually: add missing alt, convert to WebP,
fix the lazy-loaded LCP image, add width/height), and a note on expected CWV
impact. Write in the user's language.