skills/hmbown/codewhale/presentations/Gen Agent Trust Hub

presentations

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill instructions define a legitimate workflow for generating presentation files and follow best practices by requiring user consent for dependency installation.
  • [REMOTE_CODE_EXECUTION]: While the skill mentions the use of python-pptx and potential installation of dependencies, it explicitly instructs the agent to ask the user before installing anything, preventing unauthorized or silent package installation.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user input to build slide outlines.
  • Ingestion points: User-provided deck purpose and content (SKILL.md).
  • Boundary markers: None explicitly defined.
  • Capability inventory: File creation/editing via python-pptx and file conversion via LibreOffice.
  • Sanitization: None specified. However, the risk is minimal as the output consists of document files and the instructions include a verification step.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 11:56 AM
Security Audit — agent-trust-hub — presentations