What problem does it solve?
Machine translation often produces stiff, unnatural output that leaks source-language word order, breaks placeholders, and drifts in register. This Skill enforces a structured translation workflow with per-language profiles, mechanical verification checks, and a scoring rubric so translated content reads as natively written.
Core Features & Use Cases
- Context-aware translation: Infers register, domain, and terminology from existing locale files, sibling translations, and surrounding code before drafting.
- Per-language quality profiles: Ships rule sets for Korean, Japanese, Chinese, and English covering typography, register systems, and translation-ese patterns, with a template for adding new languages.
- Blocking verification gate: Runs mechanical checks (placeholder integrity, structure parity, em-dash and quote rules) plus a 5-criterion rubric before any output is emitted.
- Use Case: Translate a README into Korean and Japanese in parallel, then run diff-sync mode to patch only the sections that changed after the English source is updated.
Quick Start
Translate this README section into Korean, preserving all placeholders and matching the register of the existing locale files.