The Agent Skills Directory

[SAFE]: The skill provides legitimate utilities for PDF manipulation, extraction, and form filling using standard industry libraries such as pypdf, pdfplumber, and reportlab.
[COMMAND_EXECUTION]: The skill utilizes command-line tools such as qpdf, pdftotext, and pdftk for document manipulation as described in SKILL.md. These operations are standard for PDF processing tasks and do not involve shell injection or privilege escalation.
[SAFE]: Employs runtime monkeypatching in scripts/fill_fillable_fields.py to resolve a bug in the pypdf library specifically related to selection list field formatting. This is a localized and documented bug fix.
[SAFE]: Identified an indirect prompt injection risk surface as the skill processes untrusted PDF data.
Ingestion points: PDF content is ingested via PdfReader and pdfplumber.open in scripts/extract_form_field_info.py and scripts/fill_fillable_fields.py.
Boundary markers: Absent; there are no specific delimiters or instructions to the agent to ignore instructions embedded within the extracted text.
Capability inventory: Includes file system writes (writer.write in scripts/fill_fillable_fields.py) and standard PDF CLI operations.
Sanitization: Extracted text and form data are processed without specific sanitization before being returned to the agent context.

pdf