Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external PDF files, which represents an attack surface for indirect prompt injection if a PDF contains malicious instructions intended for the agent. Ingestion occurs through libraries such as pypdf, pdfplumber, and pytesseract. The provided scripts and instructions do not demonstrate explicit boundary markers or sanitization of the extracted text before processing by the agent.
- [DYNAMIC_EXECUTION]: The skill uses runtime code modification to resolve a functional issue in a dependency. The file
scripts/fill_fillable_fields.pycontains amonkeypatch_pydpf_methodfunction that dynamically reassigns theget_inheritedmethod of thepypdf.generic.DictionaryObjectclass. This is used to fix a specific bug in selection list handling during form filling. - [COMMAND_EXECUTION]: The skill relies on external command-line utilities for several core functions. Instructions and scripts provide patterns for invoking tools like
qpdf,pdftk, and thepoppler-utilssuite (pdftotext,pdftoppm,pdfimages) to perform PDF transformations, merging, and image conversion.
Audit Metadata