interactive

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: The primary function of the skill is to identify and route the user to specific educational and reflection workflows based on a predefined catalog of local instruction files.
  • [EXTERNAL_DOWNLOADS]: The skill includes logic to display a command to the user (npx vidbyte-skills) for installing missing utility packages. This facilitates the retrieval of external software from the npm registry but is presented as a user-facing message rather than an automated execution by the agent.
  • [PROMPT_INJECTION]: The routing logic evaluates user-provided goals and conversation history, which creates an indirect prompt injection surface. The risk is minimized because the routing is restricted to a predefined list of local file paths.
  • Ingestion points: User goals, supplied materials, and conversation history.
  • Boundary markers: Absent.
  • Capability inventory: Loads and follows hardcoded local instruction files.
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 09:14 PM
Security Audit — agent-trust-hub — interactive