agentleFS
ZachArticulateVZachArticulateV/designer-pro-and-seo/skills/seo-image-audit/SKILL.md

seo-image-audit

Skill10 starsChanged 27 days agoMITView on GitHub

Audit a page's images for SEO and performance — alt-text quality (missing, linked-empty, filename, stuffed, reused), modern formats (WebP/AVIF), responsive srcset/sizes, LCP loading (lazy hero, fetchpriority), CLS-safe dimensions, real byte and pixel budgets read from the image files, and og:image — scored, with ready-to-run conversion commands. Uses DataForSEO for live image-SERP rankings when connected; otherwise audits from page fetch and parsing alone. Trigger when the user says "image seo", "image audit", "alt text", "image optimization", "convert to webp", "convert to avif", or "image performance".

The file

Files
SKILL.md

seo-image-audit

Family: seo Status: Stable

Purpose

The audit half of image SEO (split from generation because the two have different legal, asset, and API risk). Inventories a page's images and scores each on the dimensions that affect search and performance. The free path (fetch + parse) is the whole product; DataForSEO only adds live image-SERP context.

Triggers

  • "image seo" / "image audit" / "image optimization"
  • "alt text" / "image metadata"
  • "convert to webp" / "convert to avif"
  • "image rankings" / "google images"

Inputs

  • Page URL or local HTML/build path (or an image directory)
  • Optional: target keyword(s) for alt-text relevance

Steps

  1. Run the image audit — on a live page (fetched through the shared SSRF guard) or a local build, where --assets lets it read real file bytes and pixel dimensions from the image headers:
    python3 "${CLAUDE_PLUGIN_ROOT}/scripts/seo/image_audit.py" --url <URL> --human
    python3 "${CLAUDE_PLUGIN_ROOT}/scripts/seo/image_audit.py" --file page.html --url <URL> --assets dist/ --human
    Findings by type, each listing the affected images (codes, budgets and scoring: references/seo-image-audit/image-rubric.md):
    • Alt — missing, empty on a linked image (the link loses its name), file name as alt, "image of…", over-long, keyword-stuffed, reused across images.
    • Format — legacy JPEG/PNG/GIF with no WebP/AVIF alternative.
    • Responsive — no srcset/<picture>; srcset without sizes.
    • Loading — lazy-loaded hero (LCP), several or no fetchpriority=high, below-the-fold images not lazy.
    • CLS — no width+height or aspect-ratio.
    • Size (--assets) — over 500 KB, over the 200 KB hero / 250 KB budget, intrinsic pixels > 2.5× the declared width, declared ratio ≠ file ratio. Without --assets these are listed under needs_data, never guessed.
    • Filename / social — camera or hash file names; missing or relative og:image.
  2. Judge alt text in context — the script catches the mechanical failures; read the remaining alts against what each image shows on this page.
  3. Optional enrichment — if the DataForSEO extension is configured, add image SERP rankings; otherwise state, in one line, what it would add.
  4. Deliver the findings with fixes, the biggest byte/pixel wins first, and the emitted conversions (ready-to-run cwebp / avifenc / ImageMagick commands per legacy image) — the script prints them; running them needs those tools installed.

Outputs

  • Findings by type with affected images, fixes, and a 0-100 score
  • A summary of the biggest performance wins (format + size)
  • A short "what paid data would add" note on the free path

Dependencies

  • scripts/seo/image_audit.py (required) — the image audit (stdlib header parsing)
  • references/seo-image-audit/image-rubric.md (required) — codes, budgets, scoring
  • Optional: DataForSEO extension (image SERP); seo-page (page-level context)

Notes

Pairs with seo-image-gen (the generation half). Actual file conversion (WebP/AVIF) and metadata injection require Sharp/ImageMagick if you want the skill to perform the conversion rather than recommend it.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.