Skills
skills/4444j99/a-i--skills/pptx/Gen Agent Trust Hub

pptx

Pass

Audited by Gen Agent Trust Hub on Apr 19, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by processing untrusted PowerPoint (.pptx) files and extracting their text content for analysis. \n
  • Ingestion points: Presentation files are ingested via markitdown (referenced in SKILL.md), scripts/inventory.py using the python-pptx library, and ooxml/scripts/unpack.py using zipfile. \n
  • Boundary markers: Absent. There are no explicit instructions for the agent to ignore potentially malicious embedded text during the extraction or analysis process. \n
  • Capability inventory: The skill possesses significant capabilities including shell command execution via subprocess.run (calling soffice and pdftoppm in scripts/thumbnail.py and ooxml/scripts/pack.py) and browser automation via Node.js/Playwright (scripts/html2pptx.js). \n
  • Sanitization: The skill correctly uses defusedxml in ooxml/scripts/unpack.py and ooxml/scripts/pack.py to prevent XML External Entity (XXE) attacks, which is a defensive best practice. However, it does not sanitize the extracted natural language content for potential prompt injection patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 19, 2026, 03:24 AM