pptx
Pass
Audited by Gen Agent Trust Hub on Jul 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the subprocess.run method to invoke local system utilities like LibreOffice (soffice), Poppler (pdftoppm), and Git for document validation, PDF conversion, and change tracking. These operations are essential to the skill's primary functions and are implemented using argument lists rather than raw shell strings to mitigate command injection risks. Evidence: ooxml/scripts/pack.py, scripts/thumbnail.py, and ooxml/scripts/validation/redlining.py.
- [PROMPT_INJECTION]: The skill ingests and extracts content from untrusted PowerPoint files, which creates an indirect prompt injection surface. Extracted text is presented to the AI agent for analysis or editing without explicit boundary markers or sanitization. Evidence: Ingestion points via markitdown and inventory.py mentioned in SKILL.md. Capabilities associated with this surface include file system writes via replace.py and pack.py, as well as the aforementioned command execution tools.
Audit Metadata