pdf

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection as it processes and extracts text from external PDF files.
  • Ingestion points: PDF data is ingested through pypdf, pdfplumber, and pytesseract (OCR) in the SKILL.md file.
  • Boundary markers: No specific boundary markers or delimiters are provided to wrap the extracted content to prevent it from being interpreted as instructions.
  • Capability inventory: The skill facilitates file writing and the execution of external CLI utilities for document transformation.
  • Sanitization: No sanitization or validation logic is present to filter malicious payloads embedded in document text or metadata.
  • [COMMAND_EXECUTION]: The skill documentation includes examples for executing external command-line utilities to perform PDF operations.
  • Evidence: The skill provides command-line instructions for pdftotext, qpdf, pdftk, and pdfimages to merge, split, and extract data from documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 04:30 AM