ppt-workflow-builder
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access were detected. The skill uses standard, safe tools for information gathering and file creation.
- [PROMPT_INJECTION]: The skill is designed to process external documents (articles, research corpora) which is a standard surface for indirect prompt injection. However, the risk is mitigated by the structured workflow that requires the data to be transformed into intermediate JSON outlines and planning drafts before any final assets are generated, preventing raw external input from directly controlling agent behavior.
- [COMMAND_EXECUTION]: The skill generates SVG files and manages local project artifacts. These operations are restricted to the intended project workspace and do not involve executing arbitrary shell commands or privilege escalation.
Audit Metadata