The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted PDF files, creating a potential vector for indirect prompt injection. \n
Ingestion points: scripts/extract_form_structure.py, scripts/extract_form_field_info.py, and scripts/convert_pdf_to_images.py read data directly from external PDF files. \n
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the scripts or the skill's guidelines. \n
Capability inventory: The skill has the ability to write files and execute command-line tools like qpdf and pdftotext. \n
Sanitization: There is no evidence of text sanitization or filtering before the extracted data is passed to the agent context. \n- [DYNAMIC_EXECUTION]: The script scripts/fill_fillable_fields.py implements runtime monkeypatching by overriding pypdf.generic.DictionaryObject.get_inherited to modify the library's handling of inherited field attributes. \n- [COMMAND_EXECUTION]: The skill's documentation and scripts involve the execution of external binary tools including qpdf, pdftotext, pdftk, and ImageMagick for PDF and image manipulation.

pdf