Localization

Localization Skill

Goal

Produce target-language content that feels like it was written natively — not translated. Preserve meaning and intent while adapting cultural references, idioms, tone, and register for the target audience.

Distinction: Translation vs. Localization vs. Transcreation

| Mode | When to use | |------|-------------| | Translation | Technical docs, legal text, factual content — accuracy > naturalness | | Localization | UI strings, user-facing messages, product copy — natural + accurate | | Transcreation | Marketing slogans, emotional copy, humor — intent and feeling > literal meaning |

Identify which mode applies before starting.

Steps

1. Identify source and target

Confirm: source language, target locale (e.g. tr-TR, de-DE, not just German), content type (UI, marketing, technical, legal).

If the target locale isn't specified, ask.

Success criteria: You know the exact target audience (region + formality level).

2. First pass — literal translation

Translate accurately, preserving all meaning. Note every idiom, cultural reference, or humor that won't land in the target culture — flag with [ADAPT].

Success criteria: Nothing is lost in meaning; adaptation candidates are clearly marked.

3. Second pass — cultural adaptation

For each [ADAPT] flag:

• Replace idioms with target-culture equivalents (not literal translations)
• Adjust formality: Turkish has T/V distinction (sen/siz); German has du/Sie — match the brand voice
• Localize examples, measurements, date formats, currency
• Adapt humor: if it doesn't translate, replace with something that achieves the same effect

Success criteria: A native speaker would not identify the text as translated.

4. Technical localization (if applicable)

• Placeholder handling: {name}, %s, {{variable}} — preserve all placeholders exactly
• String length: UI strings often have length constraints — flag anything that grew >20%
• RTL languages: flag if target is Arabic/Hebrew and layout changes may be needed

5. Deliver with notes

Return the localized text, then add a Localization Notes section explaining:

• Major cultural adaptations made and why
• Any ambiguities where multiple valid choices existed
• Length warnings for UI strings

Success criteria: Developer/designer can implement without guessing at intent.

Why/Failure Modes

[TODO: Explain the reasoning behind this skill's approach and common failure modes to avoid.]

Standalone vs Supercharged

[TODO: Describe how this skill works on its own vs when combined with other tools/context.]

Cross-References

[TODO: Link to other relevant skills or documentation.]

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: furkangonel
• Source: furkangonel/cowrangler
• License: MIT
• Homepage: https://cowrangler.com

Install and usage instructions live in the source repository linked above.