pdf-verification-check

PDF Verification Check

Use this skill whenever your task involves generating PDF files. Do not declare task completion until you have programmatically verified that the PDFs exist, are not corrupted, and match the expected page counts.

When to Use

• Immediately after generating one or more PDF files.
• Before reporting success or moving to the next phase of a task.
• When specific page counts are required (e.g., "1-page summary", "2-page listing").

Verification Steps

1. Identify Expected Outputs: List the file paths and their expected page counts.
2. Run Verification Script: Use run_shell to execute a Python script using PyPDF2.
3. Validate Results: Ensure the script exits successfully and reports the correct page counts.
4. Remediate: If verification fails, inspect the generation logic and regenerate.

Verification Script

Use the following Python snippet with run_shell to verify PDFs. Adjust the expected_pages dictionary to match your task requirements.

import sys
from PyPDF2 import PdfReader

def verify_pdfs(files):
    """
    files: dict mapping filepath to expected_page_count (or None if any count is ok)
    """
    all_valid = True
    for path, expected_count in files.items():
        try:
            reader = PdfReader(path)
            actual_count = len(reader.pages)
            
            if expected_count is not None and actual_count != expected_count:
                print(f"FAIL: {path} has {actual_count} pages, expected {expected_count}")
                all_valid = False
            else:
                print(f"OK: {path} has {actual_count} pages")
                
        except Exception as e:
            print(f"ERROR: Cannot read {path} - {str(e)}")
            all_valid = False
            
    return 0 if all_valid else 1

# Example usage - modify this map for your specific task
files_to_check = {
    "output/listing.pdf": 2,
    "output/map.pdf": 1
}

sys.exit(verify_pdfs(files_to_check))