SEO Image Optimizer
Audits image SEO, delivery performance, metadata, and Google Images visibility, then provides concrete optimization actions.
The skill distinguishes auditing, SERP analysis, and file optimization, and requests confirmation of the source type for AI-generated images; however, it does not fully document data flows, permission boundaries, rollback, or per-operation confirmation for URL fetching, DataForSEO, credentials, external tools, and file writes, so points are deducted.
Checks, output tables, and abnormal-case handling are reasonably specific and include tool fallbacks; however, referenced scripts, MCP calls, and key execution paths are not supplied here, and browser-support claims are inconsistent, so static evidence is insufficient and the score is conservatively limited.
Trigger phrases, URL and path inputs, audit dimensions, and output structures are clear for common image-SEO scenarios; non-fit boundaries, Chinese-language behavior, and mainland-China reachability are not documented, while DataForSEO and other overseas services may limit availability, so points are deducted.
The document has front matter, MIT licensing, a version, layered sections, examples, output templates, and an error-handling table; it lacks a skill-specific changelog, clear maintenance owner and update path, consolidated dependency notes, and FAQs, so points are deducted.
It offers actionable recommendations for alt text, dimensions, formats, responsive images, lazy loading, CLS, and metadata, with command templates; however, no skill-specific reproducible outputs or outcome validation are supplied, and some savings estimates and SEO-impact claims lack evidence, leaving material manual review and warranting a deduction.
The document includes a Google-owned source and distinguishes IPTC display effects from ranking effects; most thresholds, browser-support figures, SERP procedures, and recommendations lack citations, and no tests or independent reproduction evidence for this skill path are supplied, so points are deducted.
- File-optimization commands modify images and include batch overwrite behavior; confirm backups, scope, and rollback before use.
- DataForSEO SERP analysis depends on an external MCP extension not provided with the skill and must be skipped when unavailable.
- Browser-support, JPEG XL, and AI-image-labeling claims are time-sensitive and should be rechecked against current primary documentation.
- Chinese-output conventions and reliable reachability of overseas dependencies from mainland-China networks are not established.
What does this skill do, and when should you use it?
This is the image-focused skill within the claude-seo repository. It audits alt text, file weight, formats, responsive delivery, lazy loading, CLS prevention, filenames, and CDN signals. It also supports command-line image conversion, compression, responsive variants, and IPTC/XMP metadata injection. When DataForSEO MCP is available, it adds competitive Google Images analysis.
Reads image markup and checks img, picture, srcset, sizes, loading, fetchpriority, decoding, dimensions, and lazy-loading signals; evaluates file sizes against thumbnail, content-image, and hero-image thresholds; reviews WebP, AVIF, JPEG, PNG, and SVG usage; identifies native, Perfmatters, EWWW, and generic JavaScript lazy loaders; flags CLS risks, filenames, and CDN clues; uses exiftool, cwebp, ImageMagick, or FFmpeg to inspect metadata, convert formats, compress files, and create 400w, 800w, and 1200w variants; audits or injects IPTC DigitalSourceType labels for AI-generated images; and, with DataForSEO MCP, retrieves Google Images results to analyze ranking domains, alt-text patterns, format distribution, and keyword opportunities.
- An SEO consultant audits a client site for missing alt text, oversized images, format issues, and layout-shift risks.
- An e-commerce team converts product photography to WebP or AVIF and generates responsive variants.
- A frontend team investigates image-related LCP, CLS, or above-the-fold loading problems.
- A brand team standardizes Creator, Copyright, and Description metadata across an image library.
- A team using generative AI for product imagery checks or injects the required TrainedAlgorithmicMedia label.
- An SEO team with DataForSEO access researches competing images in Google Images.
What are this skill's strengths and limitations?
- Covers concrete checks for alt text, file weight, formats, responsive images, lazy loading, CLS, filenames, and CDN delivery.
- Provides explicit size thresholds, HTML patterns, and prioritized recommendations.
- Supports format conversion, compression, responsive variants, and IPTC/XMP metadata workflows.
- Distinguishes several JavaScript lazy-loading stacks instead of treating every non-native implementation as a regression.
- Can add Google Images competitive analysis through the optional DataForSEO MCP integration.
- Image SERP analysis requires DataForSEO MCP and an associated account.
- File optimization depends on local command-line utilities that may need separate installation.
- The source provides no standalone package, standalone test results, or independent cross-platform validation for this skill.
- Some judgments, such as whether an image is above the fold, depend on page context and cannot be inferred from every markup sample alone.
How do you install this skill?
No standalone installation procedure is documented for this skill; it is bundled in AgriciDaniel/claude-seo. The repository instructions install the collection with: git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git, followed by bash claude-seo/install.sh.
How do you use this skill?
In Claude Code, use image-related triggers such as “image optimization,” “alt text,” “image audit,” or “convert to webp,” optionally with a URL. The repository README lists /seo images <url> for image analysis and /seo images optimize <path> for file optimization. Before optimizing files, run which exiftool cwebp convert ffmpeg to check available tools. If DataForSEO MCP is unavailable, image SERP analysis is skipped.