soia-pkm-transform-article-ppt

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill includes a 'confidential' mode that enforces strict local isolation for sensitive projects. This mode mandates network=deny, uses absolute paths for state management, and ensures that intermediate and final files are kept outside of version-controlled (Git) directories to prevent accidental data leakage.
  • [SAFE]: The utility script scripts/media_bundle.py performs automated security checks, including a BANNED_TEXT scanner that identifies and blocks the inclusion of sensitive metadata (like notebook IDs) and local system path patterns (e.g., /Users/ or C:\Users\) in the final PPTX output.
  • [SAFE]: The skill employs SHA-256 hashing to verify the integrity of external PPTX templates before use. This prevents the execution of the workflow using modified or malicious templates.
  • [COMMAND_EXECUTION]: The skill orchestrates the use of local Python scripts and CLI tools (e.g., officecli, notebooklm) for PPTX generation and validation. These operations are governed by a manifest contract that prevents unauthorized file system access outside the designated project boundaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 08:47 AM
Security Audit — agent-trust-hub — soia-pkm-transform-article-ppt