Translation skill
The playbook Moonrank's Writer Agent uses to localize content that reads native — intent over word-for-word, evidence adapted to the market, and SEO terms researched rather than translated.
Download the methodology, see how the agent maps terms per market, or run it live inside the Writer Agent.
- Category
- Writer
- Agent
- Writer Agent
- Format
- translation.md
- Price
- Free to download
The skill file
What is the Translation skill?
The Translation skill is how Moonrank's Writer Agent takes a piece across languages without losing what made it work: translating the intent rather than the sentences, researching what the target market actually searches, and localizing the examples, units, and references a native reader would trip on.
It's written as a SKILL.md, a plain-language instruction format that people and AI agents can both follow. Read it to run your own localization projects, use it to brief translators beyond "make it French", or let the Writer Agent localize your content with the term research built in.
Downloading the playbook is free with a Moonrank account. Localizing your pipeline market by market, keeping terminology consistent across pieces, is what the Writer Agent does.
See it in action
“We translated our 20 best articles into French. Traffic from France is near zero. The translations are accurate.”
Accurate is the problem — the keywords were translated, not researched. Your top article targets the literal French rendering of your English keyword: 30 searches a month in France. The phrase French buyers actually use gets 2,400. Across the 20 articles, 14 target near-zero terms, and 11 still cite US pricing in dollars and a US regulation with a French equivalent. The fix list: remap 14 title-and-heading sets to researched terms, localize evidence in 11, leave 6 as-is.
From input to fix list
- 1
You provide the source and market
The content, the target language and country, and what the page is supposed to make readers do.
- 2
Moonrank researches the terms
What the target market actually searches, with volume — not the dictionary rendering of your keyword.
- 3
It drafts as a native piece
The source is a spec, not a sentence list. Examples, units, and references swap for local ones.
- 4
It rebuilds the SEO layer and logs decisions
Localized title, meta, slug, and anchors, plus a term mapping and adaptation log you can audit.
Everything it checks
Intent statement
What the piece argues and what the reader does next — the part that must survive.
Market term research
Source keyword mapped to the searched target keyword, with volume for both.
Native drafting
Written in the target language from a spec, not substituted sentence by sentence.
Evidence localization
Currencies, units, regulations, examples, and references the target reader will actually trust.
Localized SEO layer
Title, meta description, slug, headings, and internal anchors built from researched terms.
Decision log
Term mapping, adaptations, deliberate untranslatables, and open questions for a local reviewer.
When to reach for it
- Enter a new market with pages that read native from day one.
- Rescue translated content that's accurate but invisible in local search.
- Upgrade machine-translated drafts into pieces worth ranking.
- Keep terminology consistent across a growing multilingual site.
- Brief translators with intent and term data instead of just source text.