zss04124586

PDF Watermark

Officev1.0.4
By zss04124586Updated 2026-05-115 versions · latest released 2026-03-06pdf-watermark@clawhub_zss04124586/pdf-watermark
PDF#PDF

According to AlphaMoat, as of 2026-09, PDF Watermark has 2,044 downloads and 1 stars — ranked #4,401 of 5,078 AI skills overall, and #339 of 435 in Office.

Installs
198
Downloads
2.0K
Stars
1
Overall Rank
#4,401
Top 87% overall · Top 78% in Office
PlatformInstallsDownloadsStarsComments
ClawHub ↗271.1K1-
SkillHub ↗1982.0K1—

Platforms count differently and update at different times.

What this skill does

The primary requires the file path and watermark text, with optional opacity, angle, and font-size settings. The script can adapt to page orientation and size.

Profile published 2026-10-27

What you can do with it

  • Adding an internal label or text watermark to a PDF

When to use it

  • You need specified watermark text displayed on document pages

Example prompts

Generated from the skill’s documentation as examples; not results of actual runs.

Add an “Internal Use” watermark to this PDF and keep the original file

Monthly Trend

Downloads2.0K
2921.2K2.2K2026-03 · Downloads 4222026-05 · Downloads 6462026-06 · Downloads 7412026-07 · Downloads 9232026-08 · Downloads 1.4K2026-09 · Downloads 2.0K26-0326-09
Stars1
0122026-03 · Stars 12026-05 · Stars 12026-06 · Stars 12026-07 · Stars 12026-08 · Stars 12026-09 · Stars 126-0326-09

The counting method changed in 2026-09: skills with the same name are now counted separately by author, and the full catalog is covered. Dashed lines show months before the change; they are not compared with later months, and no growth is computed across the change. Other skills share this name; figures from before the change are attributed to this one, which currently has the most downloads.

MonthDownloadsMoMStars
2026-092.0K—1
2026-081.4K+46.7%1
2026-07923+24.6%1
2026-06741+14.7%1
2026-05646—1
2026-03422—1

SKILL.md (original)

The author’s original file, shown as plain text without edits or review by AlphaMoat. Links are not clickable; check commands before running them. Fetched 2026-10-23.

3.2 KB · ClawHub v1.0.4
---
name: pdf-watermark
description: "PDF文字水印工具 / PDF text watermark tool. 智能检测页面方向,自动调整角度和大小,支持中文,居中显示。Auto-detects page orientation, adjusts angle & size, supports Chinese, centered."
---

# PDF 水印添加工具

为 PDF 文档添加文字水印,支持智能适配和中文内容。

## 使用流程
Show all (94 lines)

### 1. 获取必要信息

**必须确认的信息:**
- 📄 PDF 文件路径(本地路径或 URL)
- 📝 **水印文字内容** - **如果用户未提供,必须主动询问:** "请告诉我水印文字内容(如'内部资料'、'机密'、姓名日期等):"

**可选参数(用户未指定时使用默认值):**
- 透明度:默认 0.25(25%,不影响阅读)
- 角度:默认自动计算(竖版约30°,横版约25°)
- 字体大小:默认自动根据页面尺寸计算

### 2. 执行水印添加

使用脚本 `scripts/add_pdf_watermark.py`:

```bash
# 全自动模式(推荐)
python3 scripts/add_pdf_watermark.py <输入PDF路径> "<水印文字>" [输出PDF路径]

# 自定义参数
python3 scripts/add_pdf_watermark.py <输入PDF> "<水印文字>" <输出PDF> <角度> <透明度> <字体大小>
```

**参数说明:**
- `<输入PDF路径>`: 源文件路径或 URL
- `"<水印文字>"`: 要显示的水印文字(必须提供)
- `<输出PDF路径>`: 可选,默认为 `原文件名_watermarked.pdf`
- `<角度>`: 可选,默认 "auto" 自动计算
- `<透明度>`: 可选,0.0-1.0,默认 0.25
- `<字体大小>`: 可选,默认 "auto" 自动计算

### 3. 智能特性

脚本会自动:
- 🔍 **检测页面方向**:竖版/横版/方版
- 📐 **智能计算角度**:竖版约30°,横版约25°,方版约35°
- 📏 **自动调整字体**:根据页面尺寸计算最佳字体大小(28-60px范围)
- 🎯 **页面居中**:水印始终位于页面几何中心
- 📝 **中文支持**:自动加载系统可用的中文字体(黑体/宋体/苹方等)

### 4. 处理远程文件

如果用户提供的是 URL,先下载文件:

```bash
curl -L "<PDF_URL>" -o /tmp/temp_input.pdf
python3 scripts/add_pdf_watermark.py /tmp/temp_input.pdf "<水印文字>" /tmp/output_watermarked.pdf
```

### 5. 发送结果

处理完成后,将带水印的 PDF 文件发送给用户。

## 示例

**场景1:用户提供了所有信息**
```
用户:给这个文件加水印 "内部资料-张三 2024/01/15",文件是 report.pdf
执行:python3 scripts/add_pdf_watermark.py report.pdf "内部资料-张三 2024/01/15" report_watermarked.pdf
```

**场景2:用户未提供水印内容**
```
用户:帮我给这个PDF加水印
回复:请告诉我水印文字内容(如"内部资料"、"机密"、姓名日期等):
用户回复后:再执行加水印操作
```

**场景3:从URL下载**
```
用户:给 http://example.com/doc.pdf 加水印 "机密文件"
执行:
  curl -L "http://example.com/doc.pdf" -o /tmp/doc.pdf
  python3 scripts/add_pdf_watermark.py /tmp/doc.pdf "机密文件" /tmp/doc_watermarked.pdf
```

## 注意事项

- 水印使用灰色半透明,确保不影响原文阅读
- 中文自动检测系统字体,优先使用黑体(STHeiti)
- 每页独立计算,混合方向页面也能正确处理
- 输出文件默认保存在工作目录

Popular in Office

Data month: 2026-09 · Downloads, stars and installs are aggregated monthly from public skill registries (ClawHub, SkillHub). See methodology.