pptx

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes system binaries through Python's subprocess.run to facilitate document conversion and comparison. Specifically, ooxml/scripts/pack.py and scripts/thumbnail.py call soffice (LibreOffice) for PPTX conversions, scripts/thumbnail.py calls pdftoppm (Poppler) to generate slide images, and ooxml/scripts/validation/redlining.py utilizes git diff for comparing document text. These operations are confined to local file processing.
  • [EXTERNAL_DOWNLOADS]: Documentation in SKILL.md lists dependencies such as markitdown, pptxgenjs, playwright, react-icons, and sharp. These are standard tools for presentation processing and UI rendering, typically sourced from official package registries.
  • [PROMPT_INJECTION]: The skill features a surface for indirect prompt injection (Category 8).
  • Ingestion points: Untrusted data is ingested into the agent context from user-provided PowerPoint files (via markitdown as noted in SKILL.md) and HTML slide source files (via Playwright in scripts/html2pptx.js).
  • Boundary markers: The processing scripts do not implement explicit delimiters or instructions to ignore potential commands embedded within the document content.
  • Capability inventory: The skill possesses capabilities for shell command execution (subprocess.run in various scripts) and browser automation (playwright in scripts/html2pptx.js).
  • Sanitization: While structural XML integrity is protected using defusedxml, no evidence of sanitization for natural language content was found before the data is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 08:53 PM
Security Audit — agent-trust-hub — pptx