Token导航 LogoToken导航

Agent Skill · 前端设计

document-polisher(文件抛光机)

document-polisher 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。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。

document-polisher 是前端设计类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为调用外部服务;许可证信息还不明确,第三方扫描已通过,按安装量在前端设计类的 10,216 个 Skill 里排第 6,944。

总安装 441周安装 18GitHub Stars 63下载 141核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:document-polisher(文件抛光机)
来源仓库:https://github.com/promptadvisers/claude-code-polished-documents-skills
仓库路径:skills/document-polisher
安装命令:
npx skills add https://github.com/promptadvisers/claude-code-polished-documents-skills --skill document-polisher
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/promptadvisers/claude-code-polished-documents-skills --skill document-polisher

这个 Skill 是做什么的

前端设计类 Skill 关注页面结构、视觉规范和组件写法,输出通常是可运行的界面代码。

将普通文档转换为品牌化专业样式的工具。内置 10 种高端品牌美学模板供选择应用。自动匹配字体、色彩与版式设计规范。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。输出符合商业出版标准的 DOCX 文档。

document-polisher 属于前端设计类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/document-polisher

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/promptadvisers/claude-code-polished-documents-…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要调用外部服务,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

document-polisher 怎么安装?

在终端执行 npx skills add https://github.com/promptadvisers/claude-code-polished-documents-skills --s…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

document-polisher 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

document-polisher 需要什么权限?

权限级别是调用外部服务。该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

document-polisher 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

document-polisher 用的人多吗?

公开数据是累计安装 441 次,近一周安装 18 次,下载 141 次,GitHub 63 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

document-polisher 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理前端设计类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

前端设计类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
frontend-design

前端设计 · Codex、Claude

8,823,360126,072只读MIT通过
vercel-composition-patterns

Vercel 组合模式 · Codex、Claude

3,858,81625,893需要联网MIT通过
vercel-react-native-skills

Vercel React native skills 前端 · Codex、Claude

2,468,66425,878只读MIT通过
adapt

adapt 搜索 · Codex、Claude

2,027,04023,125操作浏览器未明确通过
bolder

bolder 分析 · Codex、Claude

2,004,28823,130只读未明确通过
document-polisher

文件抛光机 · 本页

44163调用外部服务未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Document Polisher

Transform any document into a polished, professionally-styled masterpiece using world-class brand aesthetics.

Overview

This skill applies premium brand styling from 10 carefully curated brands to DOCX documents. Each brand has been analyzed for its typography, color palette, and design principles to create authentic-looking documents.

Brand Selection Menu

When a user wants to polish a document, ALWAYS present this selection menu:

╔══════════════════════════════════════════════════════════════════════════╗
║                        DOCUMENT POLISHER                                 ║
║                    Select Your Brand Style                               ║
╠══════════════════════════════════════════════════════════════════════════╣
║                                                                          ║
║  EDITORIAL                                                               ║
║  ─────────                                                               ║
║  1. The Economist    │ Deep navy + serif typography                      ║
║                      │ Best for: Reports, analysis, thought leadership   ║
║                                                                          ║
║  CONSULTING                                                              ║
║  ──────────                                                              ║
║  2. McKinsey         │ Sharp corners + bold blue accents                 ║
║                      │ Best for: Strategy decks, executive summaries     ║
║                                                                          ║
║  3. Deloitte         │ Teal-blue + pill-shaped elements                  ║
║                      │ Best for: Audits, assessments, formal reports     ║
║                                                                          ║
║  4. KPMG             │ Two-tone blue + condensed headings                ║
║                      │ Best for: Financial reports, compliance docs      ║
║                                                                          ║
║  TECH                                                                    ║
║  ────                                                                    ║
║  5. Stripe           │ Dark blue + purple gradients                      ║
║                      │ Best for: API docs, developer guides, specs       ║
║                                                                          ║
║  6. Apple            │ Minimalist + generous whitespace                  ║
║                      │ Best for: Product docs, user guides               ║
║                                                                          ║
║  7. IBM              │ Plex typography + Carbon design                   ║
║                      │ Best for: Technical docs, enterprise reports      ║
║                                                                          ║
║  8. Linear           │ Modern purple + precise typography                ║
║                      │ Best for: Product specs, changelogs, dev docs     ║
║                                                                          ║
║  PRODUCTIVITY                                                            ║
║  ────────────                                                            ║
║  9. Notion           │ Clean blue + subtle accents                       ║
║                      │ Best for: Wikis, project plans, documentation     ║
║                                                                          ║
║  DESIGN                                                                  ║
║  ──────────                                                              ║
║  10. Figma           │ Vibrant multi-color palette                       ║
║                      │ Best for: Creative briefs, design docs, brands    ║
║                                                                          ║
╚══════════════════════════════════════════════════════════════════════════╝

Workflow

Step 1: Ask the User

When a user wants to polish a document:

  1. Display the brand selection menu above
  2. Ask: "Which brand style would you like to apply? You can also tell me the purpose of your document and I'll recommend a style."

Step 2: Apply the Styling

IMPORTANT: Always use the Python script with python-docx. Never use direct XML/OOXML manipulation as it can corrupt documents.

Primary Method: Python Script (REQUIRED)

The script uses python-docx to safely recreate documents with brand styling. It:

  • Reads the source document
  • Creates a new document with brand styles applied
  • Copies all content with proper formatting
  • Preserves bold, italic, lists, and structure
# Ensure python-docx is installed in the environment
pip install python-docx

# Apply brand styling
python scripts/apply_brand.py <input.docx> <brand_name> <output.docx>

# Examples:
python scripts/apply_brand.py report.docx economist polished_report.docx
python scripts/apply_brand.py proposal.docx mckinsey styled_proposal.docx
python scripts/apply_brand.py guide.docx stripe tech_guide.docx
Available Brand Names:
  • economist - The Economist (editorial, serif)
  • mckinsey - McKinsey & Company (consulting, sharp)
  • deloitte - Deloitte (consulting, teal)
  • kpmg - KPMG (consulting, bold blue)
  • stripe - Stripe (tech, dark blue/purple)
  • apple - Apple (tech, minimalist)
  • ibm - IBM (tech, enterprise)
  • notion - Notion (productivity, clean)
  • linear - Linear (tech, modern purple)
  • figma - Figma (design, colorful)

What the Script Does

  1. Loads brand configuration from templates/brand-mapping.json
  2. Creates a fresh document (avoids corruption from XML editing)
  3. Applies document-level styles:

- Normal text style (font, size, color) - Heading 1, 2, 3 styles (font, size, color, spacing) - Title style - Page margins

  1. Copies content with formatting:

- Preserves headings at correct levels - Maintains bullet and numbered lists - Keeps bold, italic, underline formatting - Preserves paragraph alignment - Copies tables with styling

  1. Saves as new file (never overwrites original)

Brand Reference Files

Detailed brand guidelines are stored in the brands/ directory:

BrandReference FileKey Elements
The Economistbrands/economist.mdGeorgia serif, #1F2E7A navy, minimal borders
McKinseybrands/mckinsey.mdHelvetica Neue, #051C2C dark, sharp corners
Deloittebrands/deloitte.mdOpen Sans, #007CB0 teal, rounded elements
KPMGbrands/kpmg.mdOpen Sans, #1E49E2 blue
Stripebrands/stripe.mdHelvetica Neue, #0A2540 + #9966FF purple
Applebrands/apple.mdHelvetica Neue, #0071E3 blue, minimal
IBMbrands/ibm.mdArial, #0F62FE blue, sharp corners
Notionbrands/notion.mdArial, #0075DE blue, block-based
Linearbrands/linear.mdArial, #5E6AD2 purple, modern
Figmabrands/figma.mdArial, multi-color palette

Style Mapping

The templates/brand-mapping.json file contains structured data for programmatic styling:

{
  "brands": {
    "economist": {
      "name": "The Economist",
      "colors": {
        "primary": "#1F2E7A",
        "accent": "#E3120B",
        "textPrimary": "#0D0D0D"
      },
      "typography": {
        "headingFont": "Georgia",
        "bodyFont": "Georgia"
      },
      "styles": {
        "h1": { "size": 28, "color": "#1F2E7A", "bold": true },
        "h2": { "size": 22, "color": "#1F2E7A", "bold": true },
        "body": { "size": 11, "color": "#0D0D0D" }
      }
    }
  }
}

Examples

Example 1: Polish a Report with McKinsey Style

User: "Make my quarterly report look more professional"

  1. Show brand menu
  2. User selects "McKinsey"
  3. Apply styling:
cd /path/to/project
source venv/bin/activate  # if using virtual environment
python .claude/skills/document-polisher/scripts/apply_brand.py quarterly_report.docx mckinsey polished_quarterly_report.docx

Example 2: Apply Economist Style to a Guide

User: "Style this guide like The Economist"

python .claude/skills/document-polisher/scripts/apply_brand.py guide.docx economist guide_economist.docx

Example 3: Tech Documentation with Stripe Style

User: "Make this API doc look like Stripe's documentation"

python .claude/skills/document-polisher/scripts/apply_brand.py api_doc.docx stripe api_doc_stripe.docx

Dependencies

  • Python 3.x
  • python-docx: pip install python-docx

The script will check for python-docx and provide installation instructions if missing.

Tips for Best Results

  1. Match brand to content: Use consulting brands for business docs, tech brands for technical content
  2. Consider your audience: Economist for academics, Stripe for developers, Apple for consumers
  3. Don't over-brand: The goal is professional polish, not exact brand copying
  4. Test with PDF export: Some styling looks different in PDF vs Word
  5. Use virtual environment: Keeps dependencies isolated

Troubleshooting

"Module not found: docx"

pip install python-docx

Document looks wrong

  • Ensure the source document uses standard Word styles (Heading 1, 2, 3, Normal)
  • Check that lists use built-in List Bullet/List Number styles

Script path issues

Always run from the project root or use absolute paths:

python /full/path/to/.claude/skills/document-polisher/scripts/apply_brand.py input.docx brand output.docx

Adding New Brands

To add a new brand:

  1. Use FireCrawl to extract branding: mcp__firecrawl__firecrawl_extract_branding
  2. Create a new file in brands/<brand_name>.md
  3. Add entry to templates/brand-mapping.json with:

- name, description, category - colors: primary, accent, background, textPrimary, textSecondary - typography: headingFont, bodyFont - styles: h1, h2, h3, body with size, color, bold

  1. Test with sample documents

平台分布

Codex

34.82%
按下载量换算49

Claude

27.23%
按下载量换算38

Cursor

18.39%
按下载量换算26

Gemini CLI

10.15%
按下载量换算14

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

调用外部服务

该 Skill 可能调用第三方服务、云服务或外部模型 API,使用前需要确认账号、额度、数据发送范围和服务条款。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills