pdf

Warn

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill documentation and scripts utilize external command-line utilities such as qpdf, pdftotext, and pdftk for PDF manipulation.
  • [EXTERNAL_DOWNLOADS]: The instructions direct the user to install several well-known third-party Python packages, including pypdf, pdfplumber, and reportlab.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it processes content from untrusted PDF files. Ingestion points: Data is extracted from PDFs using pypdf and pdfplumber in scripts/extract_form_field_info.py and scripts/check_fillable_fields.py. Boundary markers: The agent is not provided with delimiters or instructions to ignore commands within the extracted PDF text. Capability inventory: The skill can read/write files and execute CLI tools. Sanitization: There is no sanitization of the text extracted from documents.
  • [PROMPT_INJECTION]: Misleading metadata is present in the LICENSE.txt file and script headers, which claim copyright by 'Anthropic, PBC', contradicting the provided author identity 'johnkmcleod9'.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 27, 2026, 11:45 PM