技能 pdf
📦
内容修订 r1 安全 ⚙️ 外部命令📁 文件系统访问🌐 网络访问
处理 PDF、表单和表格
PDF 工作通常需要提取、转换、表单填写和文档组装。此技能指导 Claude、Codex 和 Claude Code 完成实用的 PDF 工作流。
支持: Claude Codex Code(CC)
使用我的智能体安装
将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。
智能体请求
Review the Skillstore skill "pdf" from https://skillstore.io/skills/sickn33-pdf.md and its manifest at https://skillstore.io/api/skills/sickn33-pdf/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.你的智能体仍应展示其计划，并按安全策略请求必要确认。
Agent 可读资源
当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。
测试它
正在使用“pdf”。 用户要求从财务 PDF 报告中提取表格。
预期结果:
按页面列出检测到的表格、导出的电子表格文件名，以及表格结构较弱页面的备注。
正在使用“pdf”。 用户提供福利登记表和个人字段值。
预期结果:
一份已完成的 PDF 表单，以及一段简短的验证摘要，显示匹配的字段和未解决的条目。
正在使用“pdf”。 用户要求合并多个 PDF 文件包并添加水印。
预期结果:
一份已合并并加水印的 PDF，以及源文件顺序、页数和输出路径的简明记录。
安全审计
安全v5 • 2026/7/9 打开版本化报告
No prompt injection, malicious network behavior, or data-exfiltration intent was found. Most external-command and blocker findings are false positives from markdown formatting, JavaScript template literals, comments, or PDF field identifiers; the confirmed findings are expected local filesystem reads and writes used for PDF processing.
12
已扫描文件
1,909
分析行数
25
复核项
0
已忽略误报
能力复核项 (25)
这些是真实的本地能力，对此技能可能是预期行为，因此需要复核，但不会按已确认恶意行为计分。
中
Node.js fs operations
const existingPdfBytes = fs.readFileSync('input.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Node.js fs operations
fs.writeFileSync('modified.pdf', pdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Node.js fs operations
fs.writeFileSync('created.pdf', pdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Node.js fs operations
const pdf1Bytes = fs.readFileSync('doc1.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Node.js fs operations
const pdf2Bytes = fs.readFileSync('doc2.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Node.js fs operations
fs.writeFileSync('merged.pdf', mergedPdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Synchronous file operations
const existingPdfBytes = fs.readFileSync('input.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Synchronous file operations
fs.writeFileSync('modified.pdf', pdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Synchronous file operations
fs.writeFileSync('created.pdf', pdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Synchronous file operations
const pdf1Bytes = fs.readFileSync('doc1.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Synchronous file operations
const pdf2Bytes = fs.readFileSync('doc2.pdf');
The reference example reads local PDF files from disk. This is expected for PDF processing, but it is real filesystem access that could expose unintended files if paths are chosen poorly.
中
Synchronous file operations
fs.writeFileSync('merged.pdf', mergedPdfBytes);
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open("batch_merged.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open(output_file, 'w', encoding='utf-8') as f:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open("cropped.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open(f"chunk_{start_idx//chunk_size}.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open(json_output_path, "w") as f:
The shipped script writes output JSON or PDF data to a command-line path. This is intended functionality, but it is real local file write capability and can overwrite existing files.
中
Python file write/append
with open(output_pdf_path, "wb") as f:
The shipped script writes output JSON or PDF data to a command-line path. This is intended functionality, but it is real local file write capability and can overwrite existing files.
中
Python file write/append
with open(output_pdf_path, "wb") as output:
The shipped script writes output JSON or PDF data to a command-line path. This is intended functionality, but it is real local file write capability and can overwrite existing files.
中
Python file write/append
with open("merged.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open(f"page_{i+1}.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open("rotated.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open("watermarked.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
中
Python file write/append
with open("encrypted.pdf", "wb") as output:
The guide or reference example writes generated PDF, text, image, or JSON output to local paths. This is expected behavior, but it can overwrite files or create outputs in unintended locations.
低
Python glob/pattern matching
pdf_files = glob.glob(os.path.join(input_dir, "*.pdf"))
The batch-processing example enumerates PDF files in a user-supplied directory with glob. This is legitimate for PDF batching, but it can access more local files than intended if scoped poorly.
风险因素
⚙️ 外部命令 (62)
forms.md:60-73 reference.md:64-78 reference.md:186-206 reference.md:227-242 reference.md:257-258 reference.md:258-259 reference.md:259-263 scripts/check_bounding_boxes.py:6 scripts/check_bounding_boxes.py:41 scripts/check_bounding_boxes.py:43 scripts/check_bounding_boxes.py:53 scripts/check_bounding_boxes.py:66 scripts/convert_pdf_to_images.py:14 scripts/create_validation_image.py:12 scripts/extract_form_field_info.py:11 scripts/extract_form_field_info.py:38 scripts/fill_fillable_fields.py:34 scripts/fill_fillable_fields.py:37 scripts/fill_fillable_fields.py:84 scripts/fill_fillable_fields.py:86 scripts/fill_fillable_fields.py:87 scripts/fill_fillable_fields.py:88 scripts/fill_fillable_fields.py:89 scripts/fill_pdf_form_with_annotations.py:8 scripts/fill_pdf_form_with_annotations.py:31 SKILL.md:15-26 SKILL.md:26-33 SKILL.md:33-44 SKILL.md:44-47 SKILL.md:47-54 SKILL.md:54-57 SKILL.md:57-64 SKILL.md:64-67 SKILL.md:67-77 SKILL.md:77-82 SKILL.md:82-89 SKILL.md:89-92 SKILL.md:92-100 SKILL.md:100-103 SKILL.md:103-119 SKILL.md:119-124 SKILL.md:124-140 SKILL.md:140-143 SKILL.md:143-167 SKILL.md:167-172 SKILL.md:172-181 SKILL.md:181-184 SKILL.md:184-197 SKILL.md:197-200 SKILL.md:200-209 SKILL.md:209-214 SKILL.md:214-230 SKILL.md:230-233 SKILL.md:233-249 SKILL.md:249-252 SKILL.md:252-257 SKILL.md:257-260 SKILL.md:260-274 SKILL.md:274-280 SKILL.md:280-282 SKILL.md:282-283 SKILL.md:283-285
📁 文件系统访问 (25)
reference.md:59 reference.md:76 reference.md:137 reference.md:151 reference.md:152 reference.md:166 reference.md:59 reference.md:76 reference.md:137 reference.md:151 reference.md:152 reference.md:166 reference.md:488 reference.md:500 reference.md:524 reference.md:563 reference.md:474 scripts/extract_form_field_info.py:143 scripts/fill_fillable_fields.py:55 scripts/fill_pdf_form_with_annotations.py:93 SKILL.md:42 SKILL.md:52 SKILL.md:75 SKILL.md:247 SKILL.md:272
审计者: codex 查看审计历史 →
分享与引用此报告
分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。
复制报告链接
https://skillstore.io/skills/sickn33-pdf/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/sickn33-pdf?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/sickn33-pdf?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/sickn33-pdf/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/sickn33-pdf.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>学术引用 (APA · BibTeX · CFF)
APA 引用
sickn33. (2026). pdf security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/sickn33-pdf/audits/5BibTeX 引用
@techreport{sickn33-sickn33-pdf-2026,
author = {sickn33},
title = {pdf security audit report (audit version 5)},
institution = {Skillstore},
year = {2026},
number = {5},
url = {https://skillstore.io/skills/sickn33-pdf/audits/5},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "pdf security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
- name: "sickn33"
date-released: "2026-07-09"
url: "https://skillstore.io/skills/sickn33-pdf/audits/5"
identifiers:
- type: other
value: "skillstore:sickn33-pdf:audit:5"
description: "Skillstore immutable audit report identifier"
比较变体
13 个可安装变体每位作者的技能都保持为独立的可安装项。推荐变体按 Skillstore 证据排序。
为什么此变体排名第一
Skillstore 评分最高
作者 / 来源 Skillstore 评分 证据置信度 Skillstore 使用量 已更新
Skillstore 评分
评分依据 证据置信度: 高 59
架构
85
可维护性
87
内容
69
社区
83
规范符合性
你能构建什么
准备客户表单
填写结构化或扫描版 PDF 表单，并在最终输出前验证字段位置。
提取文档数据
从 PDF 集合中提取文本、表格、图像和元数据，用于审阅或分析。
组装办公文档
合并、拆分、旋转、加水印、裁剪和保护 PDF 文件包，用于日常办公工作。
试试这些提示
提取 PDF 文本
从此 PDF 中提取所有可读文本。保留页面顺序，总结关键部分，并标注任何看似扫描的页面。
填写 PDF 表单
检查此 PDF 是否有可填写字段。提取字段信息，映射我提供的值，验证它们，并创建已填写的 PDF。
验证不可填写表单的位置
将此不可填写表单转换为图像，识别输入区域，创建验证图像，检查边界框，并添加请求的文本注释。
构建批量 PDF 工作流
设计一个安全的工作流，用于处理此文件夹中的每个 PDF，提取表格，保存输出，报告失败，并避免覆盖源文件。
最佳实践
- 在源 PDF 的副本上操作，并保持原件不变。
- 为每个脚本或命令使用明确的输入和输出路径。
- 在向不可填写表单添加注释之前检查验证图像。
避免
- 不要在未检查文件列表的情况下，对范围过大的目录运行批量操作。
- 不要用生成的输出文件覆盖源 PDF。
- 对于重要文档，不要在未经视觉检查的情况下信任提取的表单字段。
常见问题
此技能可以填写 PDF 表单吗？
可以。它支持可填写字段，也支持针对不可填写表单的基于注释的工作流。
它可以从 PDF 中提取表格吗？
可以。它包含用于表格提取和电子表格式输出的 pdfplumber 工作流。
它适用于扫描版 PDF 吗？
可以，但扫描文档需要图像转换和 OCR 工具来提取文本。
它需要外部软件吗？
某些工作流需要 Python 包、Poppler 实用工具、qpdf、pdftk 或 OCR 工具。
它可以一次处理许多 PDF 吗？
可以。参考资料包含批量处理模式，但用户应仔细限制输入文件夹。
它会通过网络发送 PDF 数据吗？
未发现证据。审查过的脚本使用本地文件，并且仅在注释中包含文档 URL。
开发者详情
作者
sickn33许可证
Proprietary. LICENSE.txt has complete terms
Skillstore 修订
r1
版本提示
作者未声明版本。
仓库
https://github.com/sickn33/antigravity-awesome-skills/tree/main/web-app/public/skills/pdf-official引用
26421118b848d9f1efc0aa169d8a7a9e7e0a877e
维护新鲜度
2026/7/18
使用量
7 次下载 · 152 次浏览