Multilingual Blog Publisher
Turn one blog topic into localized, SEO-ready language versions and hreflang assets in a single workflow.
The skill constrains generated files to the current project root and discloses the optional SEO integration and fallback behavior. However, it does not require confirmation before multi-file writes or publication-oriented output, and its data flows, sensitive-content handling, rollback, and downstream permission boundaries are incomplete, so points are deducted.
The workflow, arguments, several failure cases, and progress messages are specified in useful detail. Because this is a static-only review, the key path was not executed; there is also no dedicated test coverage for this orchestrator, and edge-case and cross-language consistency handling remain limited, so the score stays below the static ceiling.
Triggers, inputs, defaults, output layout, formats, and the localization opt-out are reasonably clear. The ISO 639-1 wording is inconsistent with regional-code examples, while limits around language count, CMS conventions, project layouts, and mainland-China reachability of external integrations are not sufficiently evidenced.
The file provides name, description, argument guidance, dependencies, version, MIT licensing, cross-references, error handling, and a command recap. Installation, ownership, and update guidance are mostly delegated to repository-level documentation; limitations and troubleshooting are incomplete, and the repository presents inconsistent test-count claims.
The objective is clear and the planned pipeline can plausibly produce directly usable source, translated, localized, and hreflang artifacts. Nevertheless, URLs remain placeholders, fallback SEO validation is only structural without the sibling skill, and static evidence cannot establish end-to-end correctness, so the score is capped conservatively.
The skill includes auditable phases, paths, failure handling, and output examples, while the repository supplies CI and test material. Those tests primarily cover shared scripts and contract coherence rather than the complete multilingual path, and there is no execution or independent corroboration in the supplied evidence, so verification remains limited.
- No commands were executed; repository CI, demos, and test-count claims are not proof of this skill's end-to-end success.
- Before running, verify language codes, output paths, URL placeholders, and CMS requirements; retain version-control history or backups for the multi-file write.
- If Claude Code, the sibling blog skills, or external SEO services are unavailable, the result may be partial or limited to structural generation.
What does this skill do, and when should you use it?
blog-multilingual is a user-invokable orchestration skill in claude-blog for international blog production. It writes the source article with blog-write, translates it through blog-translate, and by default applies cultural adaptation with blog-localize. It then creates HTML hreflang tags, a sitemap fragment, and a CMS-ready language map; optional localized JSON-LD can also be added. It fits Claude Code users who need coordinated multilingual publishing without manually connecting these steps.
Parses the topic, language codes, source language, and output format; validates ISO 639-1 codes and regional suffixes; creates multilingual/{language}/ directories inside the current project; invokes blog-write for the source article, blog-translate for each target language, and blog-localize by default to adapt brand examples, CTAs, legal references, and formality; and writes multilingual/hreflang-tags.html, multilingual/hreflang-sitemap.xml, and multilingual/hreflang-map.json. When schema is requested or enabled by frontmatter, it adds localized JSON-LD with inLanguage and translationOfWork fields.
- A marketing team wants to publish an English article in German, French, and Spanish while keeping each version in a separate project directory.
- A solo publisher needs one command that covers source drafting, translation, cultural adaptation, and international SEO metadata.
- A multilingual site needs a CMS import map containing each version's language, slug, file, title, and description.
- An editor wants translation without cultural adaptation and can use --no-localize to skip that phase.
What are this skill's strengths and limitations?
- Combines source writing, translation, and cultural adaptation in one workflow.
- Produces language-specific files plus HTML hreflang tags, an hreflang sitemap fragment, and a CMS language map.
- Supports Markdown, MDX, and HTML, including regional language codes.
- Falls back to a self-contained hreflang generator when claude-seo is unavailable.
- Requires claude-blog's blog-write, blog-translate, and blog-localize; missing blog-write is a hard error.
- Its fallback hreflang validation is structural rather than the deeper validation provided by claude-seo.
- It does not replace real website URLs automatically; publication requires updating placeholders.
- The source provides no separate test count or independent CMS validation evidence for this skill.
How do you install this skill?
The skill is stored at skills/blog-multilingual/SKILL.md in the repository. Install claude-blog according to the README, for example in Claude Code: /plugin marketplace add AgriciDaniel/claude-blog followed by /plugin install claude-blog@agricidaniel-claude-blog. The source does not document a separate installation command for blog-multilingual; it requires blog-write, blog-translate, and blog-localize from claude-blog.
How do you use this skill?
In Claude Code, run /blog multilingual <topic> --languages <codes>, for example /blog multilingual Create a guide to AI content audits --languages de,fr,es. Optional flags include --source en, --no-localize, and --format md|mdx|html. If --languages is omitted, the skill asks for the target languages. Files are written under multilingual/ in the current project, and the generated hreflang files must still be updated with real URLs.
How does this skill compare with similar options?
Unlike /blog translate, this skill starts from a topic, creates the source article, and then handles translation, cultural adaptation, and hreflang asset generation. Compared with running writing, translation, and localization commands separately, it provides a single orchestrated entry point.