pdf

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Potential for indirect prompt injection. The skill provides multiple methods for extracting text and layout information from untrusted PDF files (e.g., using pdfplumber in scripts/extract_form_structure.py and pypdf in scripts/extract_form_field_info.py). The extracted content is returned to the agent without sanitization or boundary markers that would instruct the model to ignore embedded commands. The skill's capabilities, including file system writes and command execution, increase the potential impact of such injections.
  • [COMMAND_EXECUTION]: Integration with system-level CLI tools. The instructions guide the agent to use various system utilities such as qpdf, pdftk, and ImageMagick (magick) for PDF manipulation and image processing. While these tools are central to the skill's purpose, they provide a surface for interacting with the host operating system.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:20 PM