The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the subprocess module in scripts/doc_tools_check.py to verify the installation and version of system dependencies like Pandoc and LibreOffice. These calls are restricted to a hardcoded list of legitimate utility names and use standard version flags.
[PROMPT_INJECTION]: As the skill is designed to ingest and process external document formats (PDF, DOCX, XLSX, PPTX), it is theoretically susceptible to indirect prompt injection where malicious instructions are embedded in the processed data. However, this is a standard risk for document-processing tools and is addressed by the skill's instructions to treat operations as programmatic manipulations rather than direct command execution.
Ingestion points: Document loading logic in pypdf, pdfplumber, python-docx, openpyxl, and python-pptx (e.g., in SKILL.md and references/).
Boundary markers: None explicitly enforced in the provided text extraction snippets.
Capability inventory: The skill has file system write access for output generation and can execute specific CLI tools (pandoc, soffice) via subprocess for format conversion.
Sanitization: No specific filtering or sanitization of document content is implemented in the reference scripts.

document-processing