Listing I18n

在 OpenClaw 中把中文产品 Excel 或 CSV 本地化为 Amazon 和 Shopify 的多语言 Listing。适用于翻译产品、跨境电商 Listing 本地化、Amazon listing、Shopify listing、localize products 等场景。

Security Scan

The skill's code, scripts, and runtime instructions match its stated purpose (localizing Chinese product spreadsheets for Amazon/Shopify) and request only Python/openpyxl; it does not ask for credentials or install remote code, though product text will be handled by the OpenClaw LLM so consider data sensitivity.

Purpose & Capability

Name/description align with the delivered artifacts: included Python scripts (generate_template.py, inspect_input.py, build_output.py, validate.py) implement template generation, input inspection, Excel output building and validation for Amazon/Shopify—matching the stated listing-i18n purpose. Required binary (python3) and dependency (openpyxl) are proportional.

Instruction Scope

SKILL.md instructs the agent to use the OpenClaw LLM for translation/localization and local Python scripts for auxiliary tasks. The instructions only reference local files and the provided scripts. Important: product descriptions and other input data are intended to be sent to the LLM for translation (this is expected for the skill's purpose), so you should consider privacy/confidentiality of any proprietary or PII content before invoking the skill.

Install Mechanism

No install spec is present (instruction-only skill). Scripts are included in the skill bundle and run locally. The only runtime dependency is openpyxl (installed via pip by the user), and scripts print a clear error if openpyxl is missing. No remote downloads, obfuscated code, or unusual install locations are used.

Credentials

The skill declares no environment variables, no credentials, and no config paths. The Python scripts do not read environment secrets or external tokens—only local files (Excel/CSV/JSON). This is proportionate to the stated functionality.

Persistence & Privilege

Skill flags are default (always: false). It does not request permanent inclusion or modify other skills or system-wide settings. Autonomous model invocation is allowed by platform default but not combined with any broad credential or privileged actions in this package.

Assessment

This skill appears internally consistent: the provided scripts implement the template/inspection/build/validation workflow described in SKILL.md and only require python3 + openpyxl. Before installing/using, consider: 1) Data privacy — the LLM (OpenClaw) will receive your product text for translation, so avoid sending sensitive or proprietary data unless you accept that exposure. 2) Run the Python scripts locally in a controlled environment (use a virtualenv) and install openpyxl via pip (python3 -m pip install openpyxl). 3) Review the included example JSON and outputs and always validate translations (especially for legal/regulated product claims). 4) The repository contains no network calls or hidden endpoints in the shipped scripts, but you can inspect the .py files yourself (they are short and readable) before use. If you need higher assurance, run the scripts on a sandbox machine or redact sensitive fields before passing them to the LLM.

这个 skill 运行在 OpenClaw 上：

• 翻译和本地化改写由 OpenClaw 当前会话中的 LLM 完成
• 模板生成、输入检查、Excel 输出、结果校验由本地 Python 脚本完成

不要把这个 skill 当成独立翻译程序。它是“LLM 工作流 + 本地辅助脚本”的组合。

可用脚本

• python3 generate_template.py [output.xlsx] 生成输入模板
• python3 inspect_input.py <input.xlsx|csv> [--sheet SheetName] 检查输入文件结构，列出字段和样例数据
• python3 build_output.py <translations.json> [output.xlsx] 按约定 JSON 结构生成最终 Excel
• python3 validate.py <output.xlsx> 校验输出文件是否满足平台限制

工作流

1. 用户没有产品文件

运行：

python3 generate_template.py

告诉用户填写 Products sheet 第 3 行起的数据，再继续。

2. 用户提供了产品文件

先运行：

python3 inspect_input.py <input_file>

基于脚本输出完成三件事：

1. 告诉用户检测到的列名和前 3 行样例
2. 确认以下字段映射
3. 如果缺少必填字段，要求用户补齐

必填字段：

• product_id
• brand
• product_name
• category
• specs
• selling_points
• keywords_cn
• package_includes

可选字段：

• custom_attributes
• images_note

3. 确认平台和目标市场

如果用户没有特别说明：

• 平台默认 Both
• 目标市场默认 en_US

支持的目标市场：

• en_US
• en_UK
• de_DE
• ja_JP
• es_ES
• fr_FR
• it_IT

4. 由 LLM 执行逐产品翻译

对每个产品、每个目标市场、每个平台分别生成内容。

Amazon

• title 最长 200 chars，建议不超过 150
• bullet_point_1 到 bullet_point_5 共 5 条，每条建议 150-250 chars
• description 建议 150-300 词
• backend_keywords 最长 249 bytes，小写、空格分隔、不含品牌名

Shopify

• title 最长 100 chars
• description_html 建议 100-400 词，只用基础 HTML 标签
• seo_title 最长 70 chars
• seo_description 最长 320 chars
• tags 8-15 个，逗号分隔，小写
• product_type 英文品类路径，用 > 分隔

翻译要求：

• 不是直译，要做目标市场本地化
• Amazon 偏搜索导向，Shopify 偏品牌叙事
• 避免绝对化营销词，如 best、#1、guaranteed
• custom_attributes 的 key 保持英文，value 翻译成目标语言
• 美国站优先使用美式计量单位；英国、德国、日本按本地习惯表达

5. 生成输出文件

先把翻译结果组织成 JSON，再运行：

python3 build_output.py <translations.json> [output.xlsx]

输出 workbook 需包含：

• Amazon_<lang> sheet
• Shopify_<lang> sheet
• Source_CN sheet

Amazon sheet 列顺序必须是：

product_id
brand
title
bullet_point_1
bullet_point_2
bullet_point_3
bullet_point_4
bullet_point_5
description
backend_keywords
custom_attributes
source_product_name

Shopify sheet 列顺序必须是：

product_id
brand
title
description_html
seo_title
seo_description
tags
product_type
custom_attributes
source_product_name

6. 校验并汇报

生成 Excel 后运行：

python3 validate.py <output.xlsx>

向用户汇报：

• 处理了多少个产品
• 生成了哪些平台和语言版本
• 输出文件路径
• 校验结果中有哪些 error 和 warning

translations.json 结构

build_output.py 读取如下结构。仓库里还提供了可直接参考的 translations.example.json：

{
  "sheets": [
    {
      "name": "Amazon_en_US",
      "headers": [
        "product_id",
        "brand",
        "title",
        "bullet_point_1",
        "bullet_point_2",
        "bullet_point_3",
        "bullet_point_4",
        "bullet_point_5",
        "description",
        "backend_keywords",
        "custom_attributes",
        "source_product_name"
      ],
      "rows": [
        [
          "SP-001",
          "SoundPulse",
          "SoundPulse Noise Cancelling Earbuds",
          "【NOISE CANCELLATION】...",
          "【HI-RES AUDIO】...",
          "【ALL-DAY COMFORT】...",
          "【SWEAT RESISTANCE】...",
          "【IN THE BOX】...",
          "Long description...",
          "wireless earbuds bluetooth earbuds sports earbuds",
          "color:black | connectivity:bluetooth 5.3",
          "SP-Pro 主动降噪真无线蓝牙耳机"
        ]
      ]
    },
    {
      "name": "Source_CN",
      "headers": ["product_id", "brand", "product_name"],
      "rows": [["SP-001", "SoundPulse", "SP-Pro 主动降噪真无线蓝牙耳机"]]
    }
  ]
}

注意

• 依赖：python3 和 openpyxl
• 建议每批不超过 20 个产品
• 日语和德语输出建议母语者最终复核