Genpark Multilingual Copy Skill

An intelligent localization engine that transforms single-language English product descriptions into locale-adapted marketing variants for global markets (zh-CN, ja-JP, ko-KR, fr-FR, de-DE).

Capabilities

• Locale-Specific Adaptation: Tailors marketing copy to cultural nuances and linguistic norms across 5 major global markets:
  • zh-CN (Simplified Chinese)
  • ja-JP (Japanese)
  • ko-KR (Korean)
  • fr-FR (French)
  • de-DE (German)
• Tone Customization: Adjusts taglines, sentence structures, and marketing phrasing according to requested tones:
  • humorous
  • professional (Default)
  • eco
  • data-driven
  • imaginative
• Structured Outputs: For each requested locale, outputs:
  • tagline: Engaging locale-specific slogan
  • short_desc: Adapted product description with natural sentence starters
  • cta: Culturally aligned Call-To-Action phrase
  • tone_note: Detailed explanation of tone and linguistic adjustments made
  • source_word_count: Metric tracking input description length

Skill Structure

genpark-multilingual-copy-skill/
├── skill.json                # Skill manifest schema & metadata
├── multilingual_copy_agent.py # Core MultilingualCopyClient implementation
├── example_usage.py          # Runnable example script
├── requirements.txt          # Python dependencies
└── README.md                 # Technical documentation

Quick Start

Installation

No external third-party dependencies are required. Uses Python standard libraries (typing, json).

Usage Example

Run the included example script:

python example_usage.py

Python API Usage

from multilingual_copy_agent import MultilingualCopyClient

client = MultilingualCopyClient()

result = client.generate(
    product_name="RoboClean X10 Ultra",
    english_desc="Next-generation smart robot vacuum featuring 6000Pa laser navigation.",
    tone="professional",
    locales=["zh-CN", "fr-FR"]
)

for variant in result["variants"]:
    print(f"Locale: {variant['locale']}")
    print(f"Tagline: {variant['tagline']}")
    print(f"CTA: {variant['cta']}\n")

Input & Output Schema

Inputs (skill.json)

Parameter	Type	Description	Default
product_name	String	Name of product	Required
english_desc	String	English source description	Required
tone	String	Copy tone (humorous, professional, eco, data-driven, imaginative)	"professional"
locales	Array	Subset of ["zh-CN", "ja-JP", "ko-KR", "fr-FR", "de-DE"]	All locales

Output Format

{
  "variants": [
    {
      "locale": "zh-CN",
      "tagline": " RoboClean X10 Ultra — ，",
      "short_desc": "：Next-generation smart robot vacuum featuring 6000Pa laser navigation.",
      "cta": "",
      "tone_note": "Locale 'zh-CN' adapted with 'professional' tone. ，。"
    }
  ],
  "source_word_count": 8
}

License

MIT License. Created by alphaparkinc.