Skip to content

PDF Text Extractor Workflow

Use this skill when you need to read, parse, or extract text from an external PDF document (e.g., government directives, manuals, or research papers) provided via a URL.

Step 1: Download the PDF Safely

Standard web requests often fail on strict servers. Use curl.exe with a standard browser User-Agent (-A) and follow redirects (-L).

Action: Use the run_command tool to execute: powershell curl.exe -A "Mozilla/5.0" -L "<PDF_URL>" -o "temp_document.pdf"

Step 2: Extract Text using uv and pypdf

Do not attempt to install global Python packages. Instead, use uv to run a temporary, isolated Python script with the pypdf dependency.

Action: Use the run_command tool to execute: powershell uv run --with pypdf python -c " from pypdf import PdfReader import sys try: reader = PdfReader('temp_document.pdf') text = '' for page in reader.pages: text += page.extract_text() + '\n' print(text) except Exception as e: print(f'Error reading PDF: {e}', file=sys.stderr) sys.exit(1) "

Step 3: Analyze and Clean Up

  1. Read the standard output from the command to get the extracted text.
  2. If the text is extremely long, parse it or summarize it as requested by the user.
  3. Clean up the downloaded file by running Remove-Item temp_document.pdf.

Deep State of Mind (DSOM) For My AI Protocol | Harisfazillah Jamel (LinuxMalaysia) | 2026-07-04 Standard: UK English | DBP-standard Bahasa Melayu Malaysia (Piawai) | GNU General Public License v3.0