pdf

Warn

Audited by Gen Agent Trust Hub on Sep 29, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The script scripts/fill_fillable_fields.py implements a monkeypatch_pydpf_method function that replaces pypdf.generic.DictionaryObject.get_inherited with a local implementation at runtime to address a specific bug. This represents dynamic modification of third-party library behavior within the skill environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted PDF files, creating a vulnerability surface. 1. Ingestion points: The skill uses scripts/extract_form_field_info.py and scripts/convert_pdf_to_images.py to process external PDF files. 2. Boundary markers: The structured instructions in forms.md provide a workflow but do not contain explicit commands to ignore instructions embedded within the PDF content or metadata. 3. Capability inventory: The skill has the capability to write files to the filesystem through scripts/fill_fillable_fields.py and scripts/fill_pdf_form_with_annotations.py. 4. Sanitization: No evidence of input validation or sanitization of text extracted from or written to the PDF documents was found.
  • [COMMAND_EXECUTION]: Instructions in forms.md direct the agent to execute multiple Python scripts from the scripts/ directory using shell commands (e.g., python scripts/check_fillable_fields <file.pdf>). This behavior relies on the agent platform's execution layer to prevent command injection via malicious characters in user-supplied file paths.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 29, 2026, 03:43 PM