PDF Simple Tool

PDF splitting and PDF-to-Word conversion tools implemented in Node.js.

PDF Simple Tool Skill

This skill provides two main actions for working with PDF files:

1. split_pdf – Extract a page range from a PDF into a new PDF.
2. pdf_to_word – Convert a PDF into a simple Word (.docx) document.

Install

cd skills/pdf-simple-tool/node/ 执行命令： npm install

Actions

split_pdf

When the user asks to split a PDF by page range (for example, "把这个 PDF 的第 3-5 页拆出来" / "split pages 3–5"), call the Node implementation:

• Entry: skills/pdf-simple-tool/node/index.js
• Function: splitPdf(inputPath, outputPath, fromPage, toPage)

Inputs

• inputPath (string): The absolute path to the source PDF.
• outputPath (string): The absolute path where the new PDF will be written.
• fromPage (integer): Start page (1-based).
• toPage (integer): End page (1-based, inclusive).

Example behavior:

• If the user says: "帮我把 /Users/xingxing/.openclaw/workspace/test.pdf 的第 1-3 页拆成一个 PDF 文件", map this to:
  • inputPath = "/Users/xingxing/.openclaw/workspace/test.pdf"
  • outputPath = "/Users/xingxing/.openclaw/workspace/test_p1-3.pdf"
  • fromPage = 1
  • toPage = 3

pdf_to_word

When the user asks to convert a PDF to Word (for example, "把这个 PDF 转成 Word" / "convert to docx"), call the Node implementation:

• Entry: skills/pdf-simple-tool/node/index.js
• Function: pdfToWord(inputPath, outputPath)

Inputs

• inputPath (string): The absolute path to the source PDF.
• outputPath (string): The absolute path where the Word file (.docx) will be written.

Example behavior:

• If the user says: "把 /Users/xingxing/.openclaw/workspace/test.pdf 转成 Word", map this to:
  • inputPath = "/Users/xingxing/.openclaw/workspace/test.pdf"
  • outputPath = "/Users/xingxing/.openclaw/workspace/test.docx"

Notes

• Implementation code lives in skills/pdf-simple-tool/node/index.js and uses:
  • pdf-lib for PDF manipulation.
  • pdf-parse + docx for PDF-to-Word conversion.
• You are responsible for wiring these actions into your agent so that natural language requests are converted into the appropriate function calls with the correct paths and page ranges.