Purpose
Take a piece of source content written for one market and produce a version that reads as though it were written natively for the target market — adjusting tone, examples, references, units, and cultural framing without losing the original message or brand voice. Solves the "technically translated but culturally wrong" problem that flattens engagement in international campaigns.
When to Use
Invoke this skill when the user asks to:
- Adapt existing content (web page, email, ad, landing page, social post) for a new country or region
- Fix localised content that reads awkwardly or feels literally translated
- Localise a global campaign for a specific target market
- Adjust tone, examples, or cultural references for a regional audience
- Convert currency, units, dates, and formatting for a target market while keeping the message intact
- Sense-check whether an existing piece will land culturally before launch
Inputs Required
Before starting, collect:
- Source content: The full text to be adapted, and its content type (email / ad / landing page / blog / social / product copy)
- Source market: The market the content was originally written for (e.g., US English)
- Target market: The specific country or region, and language variant (e.g., UK English, Mexican Spanish, German)
- Brand voice notes: Any tone rules, banned words, or voice attributes to preserve
- Product/offer specifics: What is being sold or promoted, including any market-specific pricing, availability, or legal constraints
- Sensitivities to avoid: Topics, humour styles, imagery, or claims that are off-limits in the target market
- Preserve list: Any elements that must stay identical (brand name, legal disclaimers, product names, trademarked slogans)
If brand voice notes or sensitivities are missing, infer conservative defaults from the target market and flag the assumptions in the change log.
Instructions
Step 1: Decode the Source Intent
Before touching a word, identify what the content is actually doing:
- Core message: The single takeaway the reader must leave with
- Primary CTA: The action the content drives toward
- Emotional register: Where the source sits on the formal ↔ casual and reserved ↔ enthusiastic axes
- Persuasion mechanics: Which levers the source pulls (social proof, urgency, authority, aspiration)
Everything adapted downstream must still deliver this intent. Localisation changes the delivery, never the destination.
Step 2: Calibrate Tone for the Target Market
Adjust the emotional register to what reads as natural — not exaggerated — in the target market:
- Directness vs. indirectness: US/Dutch copy tolerates blunt claims; UK/Japanese/German audiences often prefer understatement and hedged confidence
- Formality of address: Determine formal vs. informal second-person where the language distinguishes it (tu/vous, du/Sie), and default to the register the market expects for this content type
- Enthusiasm ceiling: Superlatives and exclamation density that feel energetic in one market read as untrustworthy in another — dial to local norms
- Humour: Only carry humour across if the style translates; replace or remove puns and wordplay that don't survive the language
Step 3: Localise Examples, References, and Idioms
Replace source-market-specific content with target-market equivalents:
- Idioms and metaphors: Never translate literally — substitute an equivalent local expression or rewrite plainly
- Cultural references: Swap sports, celebrities, holidays, institutions, and pop culture for locally resonant equivalents
- Scenarios and personas: Adjust names, settings, professions, and daily-life details so examples feel lived-in for the reader
- Social proof: Where possible, reframe stats or testimonials around local relevance ("trusted by X companies" → a locally meaningful framing)
Step 4: Convert Technical and Formatting Elements
Apply the mechanical layer:
- Currency: Convert amounts and use the correct symbol, placement, and separators (e.g., €1.299,00 vs. $1,299.00)
- Units: Metric/imperial per market; temperature, distance, weight
- Date and time: Local date order and 12/24-hour convention
- Numbers and punctuation: Decimal/thousands separators, quotation mark style, spacing rules
- Compliance: Flag any claim, guarantee, or offer term that may require adjustment for local regulation (leave the legal decision to the operator, but surface it)
Step 5: Rebuild the Call to Action
The CTA is where localisation most often breaks:
- Ensure the CTA verb and phrasing match how the target market naturally expresses that action
- Adjust urgency framing to local tolerance
- Confirm the offer described (price, availability, delivery, guarantee) is accurate for the target market or flag it for the operator
Step 6: Produce the Change Log
Document every non-trivial adaptation so the operator can review and trust the output. Group changes by category (tone, references, formatting, CTA, flagged risks) with a one-line rationale each.
Output Format
- Structure: Two parts — (1) the fully localised content, ready to use; (2) a change log table
- Length: Localised content matches the source length; change log is 8–20 line items depending on source complexity
- Format: Localised content as clean prose/copy in the target language variant; change log as a table with columns: Element | Original | Adapted | Reason
- Sections: Localised Content → Change Log → Flagged for Operator Review (any compliance, pricing, or availability items needing a human decision)
- Tone: The localised content reads natively for the target market; the change log is plain, factual, and reviewer-friendly
Quality Checklist
Before delivering output, verify:
- [ ] The core message and primary CTA from the source are fully preserved
- [ ] No idiom or metaphor has been translated literally — every one is substituted or rewritten
- [ ] All currency, units, dates, and number formatting match the target market convention
- [ ] Every item on the operator's "preserve list" appears unchanged
- [ ] The change log has one entry per non-trivial adaptation, each with a reason
- [ ] Any compliance, pricing, or availability uncertainty is surfaced under "Flagged for Operator Review" rather than silently assumed
Notes
- Localisation is adaptation, not translation — a good result often diverges word-for-word from the source while landing the same intent more effectively
- The most common failure mode is over-enthusiasm carried across from US-style source copy into reserved markets; when in doubt, dial tone down and note it
- When the target language uses formal/informal address, choosing the wrong register is the single fastest way to sound foreign — decide it explicitly in Step 2 and hold it consistent throughout
- Flagged compliance and pricing items are deliberately left for the operator; never invent local legal terms or prices to fill a gap
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills