The Agent Skills Directory

PROMPT_INJECTION (HIGH): The skill exhibits a significant Indirect Prompt Injection surface (Category 8) because it processes untrusted external PDF documents and has side-effect capabilities.
Ingestion points: Untrusted data enters the context via PDF files read by pypdf, pdfplumber, and pytesseract.
Boundary markers: There are no boundary markers or instructions to ignore embedded commands within the PDF content.
Capability inventory: The skill possesses extensive capabilities including filesystem writes (PdfWriter), external command execution (qpdf, pdftk, poppler-utils), and complex data transformation.
Sanitization: No sanitization or validation of the extracted PDF content is performed before it is used to drive agent reasoning or form-filling logic.
COMMAND_EXECUTION (MEDIUM): The skill utilizes dynamic code execution techniques (Category 10).
Evidence: The script scripts/fill_fillable_fields.py performs runtime monkeypatching of the pypdf library, overwriting DictionaryObject.get_inherited with a custom implementation.
COMMAND_EXECUTION (LOW): The skill documentation explicitly provides examples for executing various system-level PDF utilities.
Evidence: SKILL.md contains shell snippets for qpdf, pdftotext, and pdftk to perform document modifications.

pdf