Starter Web skill
Check a page title and description
Check a web page title and meta description for unique intent, accurate content, useful context, and rendered implementation.
Describe the actual page and its primary intent.
When to use
- Use for individual pages, template families, migrations, or SEO quality review.
- Do not promise rankings, invent demand, stuff variants, or write claims the page cannot support.
Procedure
- Identify page purpose, audience, target intent, primary entity, and actual content.
- Fetch rendered HTML and record title, description, canonical, heading, status, and indexability.
- Write a concise unique title with the distinguishing subject first and brand only where useful.
- Write a natural description that sets accurate expectations and avoids unsupported superlatives.
- Compare across the template set for duplicates, boilerplate, missing entities, truncation risk, and encoding.
- Verify server output and client hydration do not conflict.
- Re-crawl representative pages and preserve before-after evidence.
--- name: check-a-page-title-and-description category: web description: Check a web page title and meta description for unique intent, accurate content, useful context, and rendered implementation. Use when reviewing search metadata or page templates. --- # check-a-page-title-and-description Describe the actual page and its primary intent. ## When to use - Use for individual pages, template families, migrations, or SEO quality review. - Do not promise rankings, invent demand, stuff variants, or write claims the page cannot support. ## Procedure 1. Identify page purpose, audience, target intent, primary entity, and actual content. 2. Fetch rendered HTML and record title, description, canonical, heading, status, and indexability. 3. Write a concise unique title with the distinguishing subject first and brand only where useful. 4. Write a natural description that sets accurate expectations and avoids unsupported superlatives. 5. Compare across the template set for duplicates, boilerplate, missing entities, truncation risk, and encoding. 6. Verify server output and client hydration do not conflict. 7. Re-crawl representative pages and preserve before-after evidence. ## Done - A metadata report records page intent, title, description, heading, canonical, uniqueness, and content alignment - Rendered and source HTML are checked across representative templates with no missing or conflicting metadata