SEO Landing

What SEO Landing does

The skill walks an agent through eight steps: analyze the existing page, fix technical SEO issues, optimize the HTML structure, improve Core Web Vitals, optimize images and fonts, add structured data, remove unnecessary dependencies, and validate the result. The target output is static HTML with critical CSS, AVIF images, full schema.org markup and zero third-party requests on first load, aimed at a 100/100 lab PageSpeed score.

Enforced requirements cover performance (srcset and sizes, width and height attributes, lazy loading only below the fold), SEO tags and JSON-LD types such as WebSite, Organization, BreadcrumbList, FAQPage and VideoObject, accessibility to WCAG 2.1 AA, security headers, and system fonts only. The repository holds the SKILL.md workflow, reference files for the technical spec, server configuration and video and map facades, a benchmark disclosure with raw Lighthouse reports, and a negative test fixture. A Russian README is included.

Key features

• Eight-step workflow from page analysis to final validation
• Static HTML with critical CSS and AVIF images
• Full schema.org JSON-LD, emitted only when backed by facts
• Accessibility target of WCAG 2.1 AA
• Server configuration reference for caching, compression and security headers on Nginx and Apache
• Benchmark disclosure with raw Lighthouse reports

When to use it

• Rebuild an AI-generated landing page that ships too much JavaScript
• Add structured data and metadata to an existing landing page
• Set up caching and security headers for an Nginx or Apache site

Who it is for: Developers and marketers who build landing pages with an AI coding agent and want technical SEO handled consistently.

How it fits with Hermes Agent

Hermes is named among the agents that can use this Agent Skills format skill, and the repository carries the hermes-skill topic; it is not specific to Hermes.

Note: The README says the PageSpeed and Core Web Vitals figures are optimization targets, not guaranteed outcomes.

FAQ