The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/fill_fillable_fields.py implements a monkeypatching technique to dynamically override the get_inherited method of the pypdf.generic.DictionaryObject class at runtime. While the skill indicates this is a fix for a specific library bug, runtime modification of loaded code represents a form of self-modifying code behavior.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from PDF documents via methods like text extraction, OCR, and image analysis, which creates a surface for malicious content to influence agent instructions.
Ingestion points: Text extraction in SKILL.md using pypdf and pdfplumber, form metadata extraction in scripts/extract_form_field_info.py, and OCR in SKILL.md using pytesseract.
Boundary markers: Absent. The instructions do not define delimiters or specify that the agent should ignore instructions embedded within the processed document text.
Capability inventory: The skill possesses capabilities for file-writing (pypdf, reportlab) and command execution (qpdf, pdftotext).
Sanitization: There is no evidence of sanitization or filtering of the content extracted from external PDFs.
[COMMAND_EXECUTION]: The documentation (SKILL.md and reference.md) encourages the execution of various command-line utilities such as qpdf, pdftotext, pdftk, and pdftoppm for rendering and manipulating documents.

pdf