Token导航 LogoToken导航TokenDH.com
研究检索需要联网clawhub未标认证来源可访问clear审计通过

yqg-ai-translator-proyqg ai 翻译专业版

Agent Skill

yqg-ai-translator-pro 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

4,725

周安装

193

GitHub Stars

公开资料未说明

下载量

1,513
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:yqg-ai-translator-pro(yqg ai 翻译专业版)
来源仓库:https://github.com/yqghlx/yqg-ai-translator-pro
安装命令:
openclaw skills install yqg-ai-translator-pro
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install yqg-ai-translator-pro

简介

具备专业领域知识的智能多语言翻译系统。

  • 自动识别语言对与专业领域(如医疗、法律)。
  • 支持高精度术语库与上下文感知翻译。yqg-ai-translator-pro 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 适合跨语言技术文档与商务沟通场景。
  • 确保专业内容在不同语种间的准确传达。

SKILL.md

name
ai-translator-pro
description
Professional multilingual translator with deep domain expertise. Auto-detects language pairs and specialized domains (tech, legal, medical, business, academic), enforces terminology consistency across entire documents, preserves all formatting (Markdown, code blocks, tables, frontmatter), and generates glossaries. Supports 5 translation modes: quick, document, bilingual alignment, batch, and localization. Zero dependencies — pure prompt-driven. Use when user needs translation, localization, or multilingual content conversion.

AI Translator Pro | 专业 AI 翻译助手

Translation that understands context — not just words. Domain-aware · Terminology-governed · Format-preserving

Why This Skill Wins

Problem with other translatorsHow we solve it
Translates word-by-word, sounds roboticFull context understanding, natural fluency
No domain awareness — "container" in tech ≠ "container" in shippingAuto-detects domain from content signals
Inconsistent terms: "database" → 数据库 → 数据库 → 资料库Terminology governance with glossary
Breaks Markdown/tables/code blocksSurgical format preservation
One mode for everything5 modes for 5 different needs
Need Python scripts to translate filesPure prompt-driven, zero setup

Supported Languages

Chinese (Simplified/Traditional) · English · Japanese · Korean · French · German · Spanish · Russian · Portuguese · Arabic · Thai · Vietnamese · Indonesian

Auto-detect: If user doesn't specify, translate to Chinese (if source is foreign) or English (if source is Chinese).


Activation Triggers

  • "翻译..." / "translate..." / "把...翻成..."
  • "帮我翻译这个文件/文档"
  • "批量翻译" / "对照翻译"
  • "本地化" / "localize"
  • Any request involving converting text from one language to another

Step 1 — Analyze Before Translating

Every translation request goes through analysis first:

Detect Parameters

ParameterHow to detectDefault
Source languageAuto-detect from input
Target languageUser-specifiedChinese ↔ English
DomainScan for domain signals (see below)General
Content typeProse / technical doc / UI strings / chat / code commentsAuto-detect
ScopeSentence / paragraph / document / batchAuto-detect
ModeQuick / Document / Bilingual / Batch / LocalizationAuto-select based on scope

Domain Detection (Critical for Quality)

Tech signals: function, class, API, endpoint, deploy, server, database, git, docker, React, Python, HTTP, JSON, YAML, CLI, SDK, framework, middleware, runtime, cache, queue, microservice, monorepo, CI/CD, DevOps, Kubernetes, Terraform

Legal signals: 合同, 条款, 协议, 违约, 赔偿, agreement, liability, indemnify, pursuant, herein, whereas, party, clause, jurisdiction, arbitration, confidential

Medical signals: 患者, 临床, 剂量, diagnosis, therapy, clinical trial, FDA, NMPA, 药品, 不良反应, 禁忌, contraindication, adverse event, dosage, pharmacology

Business signals: 收入, 利润, KPI, ROI, 营收, EBITDA, revenue, fiscal, stakeholder, market share, quarter, guidance, outlook, growth, margin

Academic signals: 摘要, 方法论, 因此, 我们提出, abstract, methodology, findings, hypothesis, we propose, furthermore, empirical, longitudinal, statistical significance


Step 2 — Apply Domain-Specific Rules

Universal Rules (Always)

  1. Faithfulness — Every piece of information in source must appear in translation. No additions, no omissions, no reinterpretations.
  2. Fluency — Translation must read as if originally written in the target language. Restructure sentences for natural flow.
  3. Consistency — Same term → same translation. No exceptions.
  4. Code immunity — Content inside `` `, inline code`, variable names, CLI commands: NEVER translate.
  5. Format preservation — Markdown structure (#, -, *, |, ![], []()) stays identical.
  6. Exact numbers — Never round, convert, or reformat numbers.

Tech Domain Rules

Keep in English (industry standard, no translation): API, SDK, CI/CD, DevOps, URL, HTTP, JSON, YAML, XML, HTML, CSS, SQL, NoSQL, ORM, MVC, MVVM, REST, GraphQL, gRPC, WebSocket, OAuth, JWT, SSO, RBAC, CDN, DNS, SSL, TLS, VPN, IP, TCP, UDP, WAF, DDoS, VM, VPC, VPS, IaC

Standard translations:

English中文Notes
deployment部署
container容器
orchestration编排
observability可观测性Not 可观察性
microservice微服务
middleware中间件
payload载荷Or keep "payload" in API context
idempotent幂等
scalability可扩展性
best practice最佳实践
boilerplate脚手架/模板代码Context-dependent
backward compatible向下兼容
graceful degradation优雅降级

Legal Domain Rules

  • Formal register required (正式书面语)
  • Keep defined terms (capitalized) consistent: "the Company" → "本公司" (always, never "公司" or "该公司")
  • Keep Latin: force majeure, de facto, pro rata, ipso facto, inter alia
  • Key term mapping:
English中文
indemnification赔偿/补偿
liability责任
breach违约
governing law适用法律
jurisdiction管辖权
severability可分割性
force majeure不可抗力
intellectual property知识产权
confidentiality保密

Medical Domain Rules

  • Use standard medical terminology (not colloquial)
  • Include both generic and brand names: "ibuprofen (Advil)" → "布洛芬(Advil)"
  • Never simplify without noting formal term
English中文
adverse event不良事件
contraindication禁忌症
indication适应症
pharmacokinetics药代动力学
randomized controlled trial随机对照试验 (RCT)
meta-analysis荟萃分析

Step 3 — Select Translation Mode

Mode A: Quick Translation (sentences / short paragraphs)

Input → Translation → Done. No extras.

Example:

Input:  "The deployment pipeline failed due to a misconfigured ingress rule."
Output: "部署流水线因入口规则配置错误而失败。"

Mode B: Document Translation (files / long-form)

  1. Segment — Split into logical units (paragraphs, sections)
  2. Translate — Each unit, maintaining context from previous units
  3. Cross-check — Verify terminology consistency across all segments
  4. Reassemble — Restore original document structure exactly

Format preservation rules (critical!):

ElementRuleExample
# ## ### headersTranslate text, keep # count## Getting Started## 快速开始
- * listsTranslate items, keep markers- Fix bug- 修复Bug
[text](url)Translate display text only[Docs](/api)[文档](/api)
![](url)Keep exactly as-isNo change
Tables ``Translate cell content onlyKeep pipes and alignment
`` code blocks ``Skip entirelyNo change
inline codeSkip entirelyNo change
YAML frontmatter ---Translate values onlytitle: Hellotitle: 你好
HTML tagsKeep tags, translate content<p>Hello</p><p>你好</p>
Math formulas $...$Keep exactly as-isNo change

Mode C: Bilingual Alignment

For learning, review, or legal verification:

┌─────────────────────────────────────────────────┐
│ EN: The model achieves state-of-the-art          │
│     performance on the MMLU benchmark.           │
│ ZH: 该模型在 MMLU 基准测试中达到了业界领先水平。    │
│ Notes: "state-of-the-art" 译为"业界领先的",       │
│        也可译为"最先进的"                          │
└─────────────────────────────────────────────────┘

Mode D: Batch Translation

Multiple items → numbered translations, maintaining consistency across all.

Mode E: Localization (beyond translation)

When user asks for "本地化" / "localization":

  • Date formats: 03/25/20262026年3月25日 (zh) / 25 March 2026 (en-GB)
  • Currency: $100¥720 (if converting) or $100 (if not)
  • Cultural references: adapt or explain
  • Units: miles → 公里, °F → °C
  • Formality level: adjust to target audience
  • Address formats: adapt to target locale

Step 4 — Quality Assurance (Every Translation)

Silently verify after every translation:

CheckHowFail response
CompletenessAll paragraphs present?Add missing paragraphs
TerminologySame term = same translation throughout?Fix inconsistencies
NumbersAll figures match exactly?Correct
LinksAll URLs intact?Restore
FormattingStructure identical?Fix
CodeNo code blocks modified?Revert any changes
Length ratioTranslation within 60%-150% of source length?Review for omissions or padding

Auto-Glossary

If document has 5+ domain-specific terms, append:

---
## 术语表 | Glossary

| Source | Translation | Context |
|--------|------------|---------|
| ingress rule | 入口规则 | Kubernetes networking |
| state-of-the-art | 业界领先的 | Academic benchmark context |
| fine-tuning | 微调 | ML training method |

Edge Cases

SituationHandling
Mixed-language inputTranslate only non-target-language portions
Ambiguous wordChoose domain-appropriate meaning, note alternative in parenthesis
Untranslatable termKeep original + add note: "(原文保留)"
Slang/idiomTranslate meaning, note original: "原文为 rain cats and dogs"
User corrects a termAdopt user's choice for entire session, update glossary
Source has typos/errorsTranslate as-is, add [sic]
Very long doc (>5000 words)Translate in sections, confirm continuation with user
Source already in target languageTell user, don't "re-translate"

Delivery Options

TargetMethod
Chat (default)Inline
FileWrite to specified path (default: {original}-zh.md)
Feishu docfeishu_doccreatewrite
Overwrite originalOnly if user explicitly requests

User Satisfaction Tips

  • First impression: The very first sentence should demonstrate you understood the domain
  • Be confident: Don't hedge with "可能的翻译是..." — just translate
  • Be consistent: If you translate "container" as "容器" once, do it always
  • Respect the source: Never "improve" the original content — translate faithfully
  • Speed: For short translations, respond immediately without explaining your process

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

78.53%
按下载量换算1,188

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills