What does this skill do, and when should you use it?
Blog Localize performs cultural adaptation after a base translation, going beyond language conversion to fit a specific market. It audits brands, statistics, calls to action, idioms, legal references, currency, cultural references, and formality, then applies locale-aware substitutions and rewrites. Built-in profiles cover DACH, Francophone, Hispanic, and Japanese markets, with a custom-locale template for other targets. It is a fit for teams that need multilingual blog content to feel locally authored rather than translated from English.
Reads a translated blog post, parses full or language-only locale codes, and loads the shared cultural-adaptation profiles. It produces an audit listing adaptation targets by severity. It can use WebSearch to find local brands, case studies, and statistics; replace foreign examples and sources; adapt CTAs, tone, forms of address, legal references, cultural references, currency, and pricing context; and preserve the original argument, source attribution, and SEO elements. It then saves the localized version, overwriting the translated file by default unless a separate localized filename is requested, and reports adaptation counts and cultural-fit scores.
- A content team has translated an English post into German and needs DACH-specific brands, DSGVO context, and consistent Sie/du usage.
- A marketing agency is preparing French content and needs to replace US examples, statistics, CTAs, and formality for a Francophone audience.
- A cross-border publishing team is adapting posts for Spain, Japan, or another market with local cultural references, currency, and examples.
- An editor finds that an existing translation still reads as translated English and wants a structured cultural audit and rewrite.
- A multilingual pipeline needs a dedicated market-adaptation step immediately after blog-translate.
How do you install this skill?
- Before use, confirm WebSearch availability, permissions, and data boundaries; treat retrieved material as unverified evidence rather than trusting model-selected local statistics or legal mappings.
- The default behavior overwrites the translated file. Use a copy or separate output path and obtain confirmation before saving when preservation matters.
- The referenced ../blog-translate/references/cultural-adaptation.md file was not included in the supplied evidence; without it, the workflow depends on the custom-locale template.
- For mainland-China users, separately verify search reachability, zh-CN cultural coverage, local legal references, and statistical sources.
- Network access
- Local filesystem
This skill is distributed as part of the AgriciDaniel/claude-blog collection at skills/blog-localize/SKILL.md. The README documents public installation in Claude Code: run /plugin marketplace add AgriciDaniel/claude-blog, then /plugin install claude-blog@agricidaniel-claude-blog; restart Claude Code to activate it. A standalone installation procedure for this individual skill is not documented in the source.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/AgriciDaniel/claude-blog.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/blog-localize" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
First run /blog translate <file> --to <code>, then run /blog localize <file> --locale <locale-code>, for example /blog localize article.md --locale de-DE. Natural-language triggers include localize blog, adapt for Germany, localiser, and adaptar. Full locale codes such as de-DE, fr-CA, and es-MX, or language-only codes such as de and fr, are accepted; ambiguous codes such as pt require clarification. The translated file is overwritten by default unless a separate output is requested.
What are this skill's strengths and limitations?
- Addresses brands, statistics, CTAs, idioms, legal references, cultural references, currency, and tone in one workflow.
- Includes profiles for DACH, Francophone, Hispanic, and Japanese markets plus a custom-locale template.
- Preserves the argument, source attribution, and SEO elements while checking consistency of formal or informal address.
- Provides a severity-based audit, quality checks, and cultural-fit scoring.
- MIT-licensed as part of the public claude-blog collection.
- Requires blog-translate to complete first; it is not a standalone base-translation skill.
- Local examples and statistics depend on WebSearch availability and the quality of discoverable sources.
- Only four built-in market groups are explicitly documented; other locales require the custom template.
- The default behavior overwrites the translated file, so preserving the original requires an explicit request.
- The source provides no independent test count or per-locale performance evidence for this skill.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Blog Localize: Cultural Adaptation for Blogs this page | 43 · Not recommended | ★ 2.4k | 14d ago | MIT |
| Multilingual Blog Publisher | 55 · Use with care | ★ 2.4k | 14d ago | MIT |
| Blog Translate: SEO Blog Localization | 48 · Use with care | ★ 2.4k | 14d ago | MIT |
| Blog Locale Audit | 50 · Use with care | ★ 2.4k | 14d ago | MIT |
| HumanAI — Text Humanization Engine | 52 · Use with care | ★ 38 | 3mo ago | MIT |
blog-translate and blog-localize are sequential rather than interchangeable: blog-translate handles the base translation and format preservation, while blog-localize handles cultural, market, compliance, and tone adaptation. The README describes /blog multilingual as combining writing, translation, cultural adaptation, and hreflang output in one pipeline.
How did FollowSkills review this skill?
The skill scopes itself to post-translation cultural adaptation and instructs the model to preserve statistical attribution. However, it requires WebSearch without defining permission boundaries, external data flows, user confirmation, sensitive-data handling, or rollback, and it defaults to overwriting the translated file. These gaps reduce the score. No malware, credential theft, or clear gross-overreach red line is shown.
The workflow, input formats, quality checks, and abnormal-locale scenarios are reasonably explicit, including missing profiles, wrong-language input, unavailable local statistics, and ambiguous locale codes. Reliability is reduced because it depends on an adjacent reference file and external search, with no skill-specific tests or static evidence reproducing key paths.
Audience, trigger phrases, locale formats, built-in DACH/Francophone/Hispanic/Japanese profiles, and custom-locale handling are clear. Non-fit boundaries and trigger-conflict rules are limited. Chinese support is only exemplified with zh-TW, and the core local-source workflow depends on WebSearch with no stated mainland-China reachability, so points are deducted.
Frontmatter, MIT licensing, author, version, argument hint, prerequisite skill, cross-references, error table, and output template are present. The selected skill lacks an independent changelog, explicit maintenance owner or update path, and complete installation guidance for its referenced file and external search dependency, so the score is reduced.
The skill addresses the main adaptation dimensions—brands, statistics, CTAs, tone, legal references, and cultural references—and specifies an audit report, save behavior, and completion summary. Correctness of local statistics, legal substitutions, and brand choices still depends on model-driven search and human review; default overwrite also creates rework risk. Static evidence cannot establish directly publishable results.
The skill supplies an auditable adaptation checklist, quality criteria, source-preservation rule, and attribution guidance. Repository material indicates CI and tests exist, but no skill-specific executed evidence, representative outputs, or cross-locale coverage is supplied, so only limited verifiability credit is justified.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →