The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses several command-line utilities for PDF manipulation, including pdftotext, qpdf, pdftk, pdftoppm, and pdfimages. These tools are invoked to perform core tasks such as text extraction, page rotation, decryption, and image rendering from PDF files.
[COMMAND_EXECUTION]: The script scripts/fill_fillable_fields.py performs a runtime monkeypatch of the pypdf library, specifically replacing the DictionaryObject.get_inherited method. This dynamic execution is used to resolve a known bug in the library's handling of PDF selection list fields.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it is designed to ingest and process untrusted PDF documents provided by users.
Ingestion points: Untrusted data enters the agent context through PDF files processed by scripts such as extract_form_field_info.py, fill_fillable_fields.py, and pdfplumber examples in SKILL.md.
Boundary markers: The instructions do not implement explicit delimiters or 'ignore' instructions for the content extracted from the documents.
Capability inventory: The skill possesses capabilities for file system read/write operations and the execution of external subprocesses.
Sanitization: There is no evidence of sanitization or filtering of text or metadata extracted from the PDFs before it is used in subsequent operations.

pdf