blog-locale-audit

This skill audits a directory of multilingual blog content for completeness, consistency, correct hreflang tagging, meta-tag parity, and freshness. It solves the problem of international content silently drifting out of sync — missing translations, mismatched section counts, stale localized posts, and broken hreflang relationships — before those issues hurt search rankings. It builds a translation coverage matrix and emits a prioritized report with runnable fix commands.

The single-file SKILL.md defines a seven-phase workflow: discovery (grouping posts by language via subdirectories, frontmatter, or hreflang-map.json), completeness audit (coverage matrix), content parity audit (section, FAQ, image, chart, word-count-ratio, link, and frontmatter checks), SEO parity audit (title, meta description, lang attribute, canonical URL, schema inLanguage and translationOfWork, alt text, slug), hreflang audit (self-referencing, return tags, x-default, valid language codes), freshness audit (source-updated-after-translation, hash and mtime drift), and a final report that can optionally be written as HTML. Notably, the workflow is security-conscious: it resolves the target directory inside the project root, rejects symlinked directories and path traversal, and requires escaping all dynamic filenames, titles, URLs, and issue text with html.escape before rendering HTML.

The target users are multilingual content teams, SEO practitioners, and localization managers running the claude-blog suite. Use cases include verifying translation coverage before a launch, catching stale translations after a source edit, and validating hreflang and canonical correctness. It reads local content and reports findings; it optionally writes an HTML report only inside the audited project root, and takes explicit defensive measures against traversal and injection.