Use this when
multilingual blog, write in multiple languages, international blog, multiple languages.
Process
One topic, one source language (default the user's working language), and a comma-separated list of Google-compatible hreflang target codes. If 10 or more target languages are requested, stop before writing anything and ask for a reviewed batch of at most 9 — this is a scaled-content-abuse guardrail, not a technical limit.
Sequence: write the original post first (through the normal blog-write playbook, so template selection, sourced statistics, and schema priorities all still apply); translate it into every target language (can run in parallel per language); unless the user opted out, run cultural localization on every translation (also parallelizable); then generate the international SEO assets.
Assets to generate: hreflang <link> tags for every language including a self-reference and an x-default fallback (every href a fully-qualified absolute HTTPS URL, every relationship reciprocal); an equivalent hreflang sitemap fragment; a machine-readable hreflang map (JSON) for CMS integration listing every version's language, slug, URL, canonical, and whether it's the x-default; and localized Article/BlogPosting schema per version with inLanguage and a translationOfWork reference back to the source.
Report a delivery summary: the source file, a table of every translated/localized file with keyword-adaptation counts, the generated SEO assets, and explicit next steps — the placeholder URLs in the hreflang assets need real absolute URLs before publishing, and the sitemap fragment needs merging into the real sitemap.
Attribution
claude-blog credits this skill's methodology to Chris Mueller's claude-blog-multilingual submission (https://github.com/Chriss54/multilingual-int, license unverifiable — the repository is not currently publicly accessible). What's reused here is claude-blog's own MIT-licensed distributed text, the same basis as every other playbook in this team.