The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied PDF files, which represents an indirect prompt injection surface where embedded text or metadata could influence agent behavior.
Ingestion points: PDF content is ingested through libraries like pypdf, pdfplumber, and CLI tools like pdftotext (referenced in SKILL.md, forms.md, and reference.md).
Boundary markers: The instructions do not specify explicit delimiters or warnings for the agent to ignore instructions embedded within the extracted PDF text.
Capability inventory: The skill allows file creation and modification (pypdf, reportlab), image processing (PIL), and command execution (magick, pdftotext, qpdf).
Sanitization: No specific sanitization or filtering of the extracted text content is implemented before it is processed by the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to use external command-line utilities such as magick (ImageMagick), pdftotext, qpdf, and pdftk for image cropping, text extraction, and PDF manipulation.
[DYNAMIC_EXECUTION]: The script scripts/fill_fillable_fields.py uses runtime monkeypatching to modify the DictionaryObject.get_inherited method of the pypdf library. This is a targeted modification to ensure correct handling of inherited form field attributes during the filling process.

pdf