AgentSkillsCN

pdf

确保代码与设计紧跟行业最新标准、高端 UI/UX 趋势，以及各类库的官方文档。

SKILL.md
--- frontmatter
name: pdf
description: Use this skill when you need to read or process PDF files. Converts PDFs to markdown for reliable text extraction, especially for large PDFs that exceed size limits.
allowed-tools:
  - Bash
  - Read
  - Glob

PDF Reading Skill

When you need to read content from a PDF file, use this skill to convert it to markdown first, then read the markdown file.

Why Use This Skill

  • •Large PDFs often fail to load due to size limits
  • •PDF text extraction is more reliable through pdftotext
  • •Markdown output preserves layout and is easier to process
  • •Works with single files or entire directories of PDFs

How to Process PDFs

Step 1: Convert PDF to Markdown

Use the pdf_to_markdown.sh script located in this skill's directory:

bash
# Single PDF file
pdf_to_markdown.sh "/path/to/file.pdf" "/tmp/pdf_output.md"

# Multiple PDFs in a directory
pdf_to_markdown.sh "/path/to/pdf_directory/" "/tmp/combined_output.md"

Step 2: Read the Markdown Output

After conversion, read the markdown file using the Read tool:

code
Read /tmp/pdf_output.md

Step 3: Clean Up (Optional)

Remove temporary markdown files when done:

bash
rm /tmp/pdf_output.md

Workflow Example

When a user asks you to read or analyze a PDF:

  1. •First, convert the PDF to markdown using the script
  2. •Read the generated markdown file
  3. •Process/analyze the content as requested
  4. •Provide your response based on the extracted text

Requirements

The script requires pdftotext from poppler. If not installed:

  • •macOS: brew install poppler
  • •Ubuntu: sudo apt-get install poppler-utils

Notes

  • •The script preserves text layout from the original PDF
  • •UTF-8 encoding is used for proper character support
  • •Failed extractions are noted in the output file
  • •For very large PDFs, you may need to read the markdown file in chunks using offset/limit parameters