en-zh-translation-polish skill
A skill that translates and polishes English into Chinese by the Ye Zinan method, removes translationese and pairs paragraphs.
- 115 ★ on the repository
- Last commit
Minimal access
Why this level
- The skill edits text and runs only a punctuation script, while publishing decisions stay with a human.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/en-zh-translation-polishDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill edits text and runs only a punctuation script, while publishing decisions stay with a human.
This is third-party code. Review the repository files before installing.
What it does
The skill turns the method from Ye Zinan's book on English-to-Chinese translation into a sequence of steps. It first classifies the text type and translation freedom on a 1 to 10 scale: legal and technical texts are adjusted cautiously, essays and commentary more freely. It then detaches from English syntax, translates at the chosen level and runs through three reference tables: translationese symptoms, techniques and metaphor decisions. Next come rhythm polishing, an accuracy check against the source and back, Chinese punctuation normalization by script, and paragraph-by-paragraph output with a line on the chosen level and key decisions. For long texts, translation and checks run in batches. If the user asks for an abridgment or a rewrite, the skill delivers it and states the limits.
Who it is for. Translators and editors who render English articles into Chinese and want calqued constructions removed.
Good fit when
- Machine translation sounds stiff and you need to find the causes and fix the text.
- A translation needs checking for omissions, unsupported additions and lost qualifiers.
- A contract or technical document is translated as closely as possible without embellishment.
- An article in two languages is kept in one file with paragraph alignment.
Not a fit when
- You need translation from Chinese into English or other pairs: the skill covers English into Chinese only.
- The goal is a quick draft translation without checks or tables.
- You need offline AI translation without an agent: this is a skill for an agent environment.
Example request
Translate this paragraph on a supply contract into Chinese, choose the freedom level and show the English-Chinese paired output.Limitations
The skill covers only the English-to-Chinese pair, and the main documentation and reference tables are in Chinese. Python 3 is needed for punctuation normalization. Output depends on the model and on how precisely the freedom level is set, and the method rests on one book and reflects its view of translation. The skill does not replace professional review of legal or medical texts. The README version is 1.1.1.
How to disable. Delete the en-zh-translation-polish folder from the agent skills directory, for example ~/.agents/skills/, and its link in ~/.claude/skills/ if you created one.
FAQ
How do I install the skill?
Run `npx skills add -g HoraceLuBFA/en-zh-translation-polish`. Manually, clone the repository into `~/.agents/skills/en-zh-translation-polish`.
What does the freedom level from 1 to 10 mean?
The lower the level, the closer the translation stays to the original and the more restrained the editing. Higher levels allow restructuring of phrases and rhetoric.
Which reference tables does the skill use?
Three: translationese symptoms, techniques and metaphor decisions. They sit in the `reference/` folder.
Is Python required?
Yes, Python 3 is needed for the script that normalizes Chinese punctuation.
Is there English documentation?
Yes, the repository has `INSTALL.en.md`, while the main README and reference files are written in Chinese.
Discussion
No questions or reviews yet.
Related
Skillsqu-ai-wei (去 AI 味)
qu-ai-wei
A skill that edits Simplified Chinese: cuts stock phrases and calques while keeping facts, numbers and the author's voice
Human Writing
human-writing
A skill for lively Chinese writing: helps the agent write and edit so a specific person seems to speak, not a model
An editor skill for Chinese text: removes filler, repetition and template phrasing, keeping facts, certainty and author voice
speak-human-tw
說人話 speak-human-tw
A skill that proofreads Traditional Chinese: strips AI writing tells and fixes Taiwan localization
A skill that teaches an agent to write Turkish without English calques: it fixes sentence structure, collocations and UI copy
A skill for writing in Chinese: goal and facts first, then platform, tone and a self-check of the draft before delivery.
README in short
The README walks through the method, from fidelity to the source to the limits of polishing, followed by a table of eight stages. The project ships on GitHub and as a ZIP from the author's site with a SHA-256 checksum. The repository includes test prompts.
README badge
Original SKILL.md
---
name: en-zh-translation-polish
description: Translate English passages into natural Chinese or polish Chinese translations against the source; supports bilingual output.
license: MIT
allowed-tools: Read Write Edit Bash
metadata:
version: 1.1.1
tags: [translation, chinese, english-to-chinese, 英译汉, 润色, 中英对照, localization]
---
# 英译汉翻译润色 (English→Chinese Translate & Polish)
把英文译成**忠实、地道的中文**并产出**英中对照**译文。借鉴叶子南的英译汉方法，在完整、准确传达原文的前提下，**以读者为中心、按文本类型调节归化尺度、发挥汉语意合优势、兼顾音韵节奏**。
**忠实性优先，适用于所有档位。**归化自由度只调整表达方式，不授权增删事实、改变逻辑关系、说话者、限定条件或语气强度。原文的比喻、夸张和个人语气应保留其作用，不额外强化。外部知识用于消歧，不能补写成原文的陈述；原文未展开的信息不自行补齐。用户明确要求摘要、节译或改写时按该范围交付，并说明处理方式。
## 按请求选择交付与检查范围
聊天中的短段落翻译直接在回复中交付，用户要求文件时才落盘；全文或项目翻译沿用下述默认文件契约。用户明确选择纯译文、双语或已有目标文件时按其要求，不重复确认。先保留数字、术语、引文与限定条件，再按文体选择必要的润色检查；短文本不必逐项展示诊断过程或通读所有参考表。研究论文、文学和正式文件的专项边界继续适用。
## 独立使用与文档输入
本技能可独立安装使用，无需额外安装其他翻译技能。输入为 PDF 或其他文档时，先使用当前宿主可用的读取工具获取原文；保留用户要求的正文、图表说明、公式、引文和参考文献。无法可靠提取的内容应说明缺口，不将翻译任务转交给用户未安装的技能。
## 交付物
默认产出**一个**英中对照 Markdown 文件；用户要“全中文版”时再脚本派生。
1. **`<名称> 翻译(中英对照).md`** — 唯一真源。每段**英文原文作为 blockquote (`> `)**，紧跟其下是中文译文。
2. （按需）**`<名称> 翻译(全中文).md`** — 脱离对照文件**用脚本机械派生**（删掉 `> ` 行），不要重打中文。
```bash
python3 - "<名称> 翻译(中英对照).md" "<名称> 翻译(全中文).md" <<'PY'
import sys, re
src, dst = sys.argv[1], sys.argv[2]
lines = open(src, encoding='utf-8').read().split('\n')
kept = [ln for ln in lines if not ln.lstrip().startswith('>')]
open(dst,'w',encoding='utf-8').write(re.sub(r'\n{3,}','\n\n','\n'.join(kept)).strip()+'\n')
PY
```
> 顺序要求：**先跑阶段 6 的标点归一，再派生全中文版**——派生脚本原样继承标点，不会再纠正。
若输入只是聊天里的一小段（几句话），直接在回复里给对照即可，不必落盘；成篇文章或用户指明保存才写文件。命名沿用源文件名，无工具前缀（除非用户要）。目标文件已存在则更新而非新建。
## 工作流（严格按序；每段译文都要走完润色与核查）