The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the execution of Python scripts and inline commands (e.g., python -c "...") to perform document analysis, table creation, and formatting operations. This is central to its functionality.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8).
Ingestion points: The skill reads external data from Word documents using docx.Document('source.docx') as seen in SKILL.md and scripts/template.py.
Boundary markers: There are no boundary markers or delimiters used when printing the content of the document paragraphs into the agent's context, making it easier for an attacker to inject instructions.
Capability inventory: The skill utilizes file system read/write capabilities via the python-docx library and directory listing via the glob module.
Sanitization: No sanitization, filtering, or escaping is performed on the text extracted from the document before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill depends on the python-docx library. While this is a well-known and standard library for document manipulation, users should ensure it is installed from a trusted registry.

docx-perfect