Claude/ChatGPT Prompt to Add SEO Metadata and Sitemap to Next.js App Router

What this prompt does

This prompt makes the AI add a complete SEO layer to a Next.js 15 App Router project, returning copy-ready code rather than pseudocode. You set the [route], the [base_url], the [og_image], and the [schema_type]. It produces generateMetadata per route (including the dynamic [route]), OpenGraph and Twitter Card tags using your default share image, JSON-LD for [schema_type] via the Script component, a dynamic sitemap.ts, a robots.ts, and canonical URLs built from [base_url].

The structure works because SEO breaks quietly when metadata and sitemaps drift apart, and a wrong canonical splits ranking across duplicate URLs. By generating all six pieces together and tying canonicals to [base_url], the prompt keeps them consistent from the first deploy. [route] shows how a dynamic page generates its metadata, [og_image] becomes the default share image, and [schema_type] (like BlogPosting) determines the JSON-LD shape so search engines and crawlers read the right structured data.

When to use it

• You're starting an App Router build and want the SEO layer wired from day one.
• You need generateMetadata for dynamic routes with title and description fallbacks.
• You want OpenGraph and Twitter Card tags with a consistent default share image.
• You need JSON-LD structured data for a specific schema type.
• You want a dynamic sitemap.ts and a sensible robots.ts pointing at it.
• You keep getting canonical URLs wrong and splitting ranking across duplicate paths.

Example output

You get each file fenced and copy-ready with a note on where it lives in the app directory: generateMetadata per route including the dynamic [route] with title/description fallbacks; OpenGraph and Twitter Card tags using [og_image] as the default share image; JSON-LD for [schema_type] injected via the Script component; a sitemap.ts that resolves static and dynamic routes from data; a robots.ts with allow/disallow rules pointing at the sitemap; and canonical URLs built from [base_url] so duplicate paths don't split ranking.

Pro tips

• Set [base_url] to your real production domain — canonicals are built from it, and a wrong value silently splits ranking across duplicate URLs, which is the hardest issue to spot post-launch.
• Match [schema_type] to the page's actual content; BlogPosting and Product generate different JSON-LD, and the wrong type sends misleading structured data to crawlers.
• Provide a real [og_image] path that exists and is the right dimensions, since a missing share image produces broken link previews.
• Use a representative [route] so the dynamic generateMetadata pattern transfers cleanly to your other dynamic pages.
• Verify the sitemap.ts resolves your dynamic routes from the same data source your pages use, so the sitemap never lists URLs that 404.
• Validate the generated JSON-LD in a structured-data testing tool before trusting it; small schema mistakes are easy to miss by eye.