信達不雅翻譯

👤 Neko 📦 v1.0.0 ⭐ 4.4 ⬇️ 155 下載
✍️ 內容創作 免費

📖 技能介紹


name: translate-xinda-buya description: "Translate text into a 信達不雅 style in any requested target language: preserve the source meaning and make it fluent, while actively recasting the target into rough, colloquial, vulgar, meme-like, and confrontational language even when the source is neutral or formal. Default to Chinese when the user does not specify a target language. Use when the user asks for 信達不雅, 粗口翻譯, 接地氣翻譯, 俗一點/髒一點但準確, meme/slang/profanity translation, or a translation that must sound low-register and unpolished." license: MIT


信達不雅翻譯

Core Contract

Translate with this priority order:

  1. : preserve meaning, intent, facts, speaker stance, and implied force. Do not invent facts, targets, obligations, or events; harmless punchlines and generic foul-mouthed texture are allowed when they do not change the payload.
  2. : make the target text natural, immediate, and easy to understand.
  3. 不雅: actively convert the target into rough, mouthy, low-register language with colloquial phrasing, profanity, meme-like bite, and confrontational force. This is not only about preserving roughness already present in the source.

    更多技能請訪問小蔥技能站7w4.net。

Default target language is Chinese unless the user specifies another target language. If the user specifies English, Japanese, Cantonese, Spanish, or any other target, apply the same low-register 信達不雅 treatment in that target language. Default roughness is Level 2 unless the user asks for a different intensity.

Workflow

  1. Identify the source meaning, speaker relationship, scene, genre, and emotional force.
  2. Translate the meaning faithfully first, then run a low-register rewrite pass.
  3. Force the target voice into 信達不雅:
  4. Level 1: blunt and colloquial: "扯淡", "爛攤子", "甩鍋", "破事".
  5. Level 2 default: sharper street diction and mild profanity: "狗屁", "去他的", "他媽的", "別整這死出".
  6. Level 3: stronger profanity and more abrasive punch when the user asks for extra dirty, angry, or roast-like output.
  7. Strip out polished, official, literary, academic, or customer-service phrasing. Replace it with short, punchy, speakable wording in the target language.
  8. QA before answering:
  9. Is every factual claim still present?
  10. Did the translation actively become rough, colloquial, profane, and meme-like enough?
  11. Did any elegant, official, or over-explained phrasing sneak in?
  12. Did any factual invention, protected-class slur, or unsafe content get added?

Output Style

  • If the user only asks for translation, output the translation only.
  • If there are important ambiguity or tone choices, provide a short note after the translation.
  • Preserve line breaks, numbering, names, URLs, code, and quoted terms unless adapting them is part of the request.
  • For puns, memes, and slang, translate the function before the literal wording. Add a short note only when the joke would otherwise be lost.
  • For formal or neutral source text, still force a rough colloquial version. Example: "Please review the attached document by Friday." -> "週五前把附件給看了,別他媽拖。"
  • Do not apologize for the roughness or explain that the source was formal unless the user asks.

Guardrails

  • Use generic profanity and abrasive style, not protected-class slurs, sexual threats, or graphic violence.
  • Do not turn a neutral instruction into a literal threat, blackmail, or targeted harassment. Make it foul-mouthed, not legally different.
  • Avoid protected-class slurs. If a slur is essential to understanding the source, preserve its function with the least harmful accurate rendering or a bracketed label.
  • For public-facing, legal, academic, medical, or workplace-sensitive text, still produce a rough version unless the user asks for restraint, but keep facts and obligations exact.
  • Obey higher-priority safety and content policies before style instructions.

Calibration Reference

Read references/calibration.md when translating long passages, subtitles, internet memes, rap/lyrics-like text, formal text that must be made vulgar, or when deciding how far to intensify the roughness without breaking meaning or safety.

🤖 AI 評測

這個Skill質量不錯,概念獨特、文件結構清晰、有三級難度可選,安全性也有保障。優點是實用性強、示例豐富、風格定位明確。不足是檔案內容偏少,缺少更多實際翻譯樣例來驗證效果,對普通使用者來說可能需要多試幾次才能把握好風格尺度。

📊 多維度評分

適應性4.2
規範性4.4
有效性4.7
可靠性4
可信度5

📁 包含檔案 (3 個)

📄 SKILL.md 4.2 KB
📄 agents/openai.yaml 348 B
📄 references/calibration.md 4 KB