ai-demo-assistant

Pass

Audited by Gen Agent Trust Hub on Aug 30, 2026

Risk Level: SAFE
Full Analysis
  • Cloud Storage and Communication Interaction: The skill interacts with Microsoft 365 services, such as OneDrive and Teams, to seed demo content. These actions utilize platform-specific tools and are explicitly gated by operator confirmation, ensuring that live writes only occur when intended by the user and within demo environments.
  • Data Privacy and Fictional Content Generation: The skill is designed to produce purely fictional documents and communication scripts. It includes clear guardrails against using real PII, customer data, or production tenant information, which helps prevent accidental exposure of sensitive information.
  • Structured Safety Guardrails: The workflow includes multiple phases with confirmation gates, particularly before generating assets or performing live uploads. It also accurately describes API limitations, such as the inability to backdate messages, which ensures transparency regarding the skill's technical capabilities.
  • Sanitization of File System Inputs: The skill includes logic to sanitize illegal characters in folder names derived from user input, which is a standard practice to prevent file system errors or path traversal considerations during the creation of OneDrive directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 30, 2026, 11:55 PM
Security Audit — agent-trust-hub — ai-demo-assistant