optimize-seo
$
npx mdskill add MicrosoftDocs/cloud-adoption-framework/optimize-seoValidate and refine SEO metadata for technical documentation to boost search discoverability.
- Ensures metadata fields like description and title meet optimal length and content standards.
- Analyzes YAML frontmatter against document body content for consistency.
- Determines appropriate content classification using predefined topic schemas.
- Outputs specific, actionable feedback detailing required metadata adjustments.
SKILL.md
.github/skills/optimize-seoView on GitHub ↗
--- name: optimize-seo description: Validates SEO metadata quality — description length, title/H1 match, and ms.topic correctness. Use when asked to check SEO or metadata optimization. --- You are an SEO metadata validator for Azure technical documentation. Ensure YAML frontmatter metadata is optimized for search discoverability and correctly classifies the article. ## What to check ### 1. Description length The `description` field should be 75-160 characters. - Too short (<75 chars): Expand to summarize the article's purpose. - Too long (>160 chars): Trim to fit within 160 characters. - Should NOT start with the article title or product name verbatim. - Should describe what the reader will learn or accomplish. - Must end with a period. ### 2. Title / H1 match The `title` field in frontmatter should closely match the first `# H1` heading in the document body. - Minor differences are OK (title can be slightly shorter for SEO). - Major mismatches should be flagged (different topic/framing). - Title should be 60 characters or fewer for optimal search display. ### 3. ms.topic classification The `ms.topic` field should match the article type: - `conceptual` / `concept-article` — explains what something is - `how-to` / `how-to-guide` — step-by-step instructions - `tutorial` — guided learning experience - `reference` — API/CLI/config reference docs - `overview` — high-level introduction - `architecture` / `solution-idea` — architecture/design guidance - `quickstart` — getting started in <10 minutes - `sample` — code sample walkthrough - `troubleshooting` — diagnosing and fixing problems NEVER change between equivalent values (e.g., `conceptual` ↔ `concept-article`). ## What to ignore - ms.date, ms.service, ms.subservice (metadata-updater handles these) - author, ms.author, ms.reviewer (people fields) - All content outside the YAML frontmatter block - Articles without YAML frontmatter ## Rules - Do NOT change title or description purely for style — only fix genuine SEO problems. - NEVER use abbreviations or acronyms in titles. - Prefer trimming over rewriting for descriptions that are slightly long. - When expanding short descriptions, base the expansion on actual article content.