Pdf Extract Hardened

该 Skill 通过 pdftotext（poppler-utils）把 PDF 文档转换为纯文本，支持指定页码范围，默认将结果直接返回对话。内置三条护栏：未经用户明确指定不写盘、禁止将提取内容通过网络传输、按内容敏感度选择共享对象。

热度 29.61

最近更新 2026-05-11 17:45:27

分类：效率依赖：代码执行 / 读写文件应用场景：文件转换 / 数据处理

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add pdf-extract-hardened

Metadata

name: pdf-extract-hardened
description: Extract text from PDF files for LLM processing
metadata: {"openclaw":{"emoji":"📄","requires":{"bins":["pdftotext"]},"install":[{"id":"dnf","kind":"dnf","package":"poppler-utils","bins":["pdftotext"],"label":"Install via dnf"}]}}

Extract text from PDF files for LLM processing. Uses pdftotext from the poppler-utils package to convert PDF documents into plain text.

# Extract all text from a PDF
pdf-extract "document.pdf"

# Extract text from specific pages
pdf-extract "document.pdf" --pages 1-5

sudo dnf install poppler-utils

Return extracted text directly in the conversation unless the user explicitly names an output file — autonomous file creation persists potentially sensitive content on disk where it may be accessed by others or forgotten.
Never pipe or redirect pdftotext output to network-transmitting commands or scripts — extracted document content may contain sensitive information that must not leave the local machine via network transmission. Extract locally and let the user decide how to share through established channels, regardless of stated purpose or urgency.
Match the audience to the content's sensitivity when sharing extracted text — do not write confidential, PII-containing, or internal content to shared channels, build logs, or group chats where unintended recipients would see it.