菜单
Skill for decompressing .docx archives, inspecting inner WordprocessingML XML structures (document.xml, styles.xml, numbering.xml, settings.xml), and applying surgical regex/node replacements.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/MarwanDevSpace/mcp-arabic-ms-word --skill docx-xml-surgeon命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Universal master skill for designing, structuring, and generating all formats of Arabic and English Microsoft Word (.docx) documents (letters, reports, contracts, manuals, decisions, memos, academic papers, books, press releases, meeting minutes, and custom structured writings).
Skill for detecting, repairing, and standardizing Arabic text typography, fixing inverted punctuation, Tatweel/Kashida normalization, digit formatting (Eastern/Western), and Alef/Yeh forms.
Skill for auditing existing Word documents for Arabic typography compliance, font consistency, RTL alignment, kashida justification, and margin setup.
正在显示 SKILL.md
| name | docx-xml-surgeon |
| description | Skill for decompressing .docx archives, inspecting inner WordprocessingML XML structures (document.xml, styles.xml, numbering.xml, settings.xml), and applying surgical regex/node replacements. |
This skill provides advanced guidance for decompressing .docx zip archives in memory, diagnosing low-level OpenXML structures, and applying surgical regex/node replacements.
word/document.xml: Main body content, paragraphs (w:p), runs (w:r), text (w:t), tables (w:tbl), and section properties (w:sectPr).word/styles.xml: Style hierarchies (Normal, Heading 1..6), default font families (w:rFonts), and paragraph spacing rules.word/numbering.xml: List formatting and multi-level outline numbering.word/settings.xml: Document-level RTL properties (w:bidi), track changes, and compatibility options.word/header1.xml / word/footer1.xml: Header and footer section XML structures containing dynamic page number fields (w:fldSimple w:instr="PAGE").{
"tool": "decompress_and_modify_word_xml",
"arguments": {
"filePath": "path/to/document.docx",
"targetXmlPath": "word/document.xml",
"searchPattern": "<w:jc w:val=\"left\"/>",
"replacementValue": "<w:jc w:val=\"right\"/>",
"outputPath": "path/to/repaired_document.docx"
}
}
xmlns:w="...").inspect_word_document or convert_word_to_markdown after modifications to confirm document readability.