risk-metrics-calculation
Pass
Audited by Gen Agent Trust Hub on Oct 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions focus on local data analysis and reproducible workflows. It explicitly mandates that the core work remains local and that external SandBase APIs are only used for user-requested tasks.
- [SAFE]: All external API interactions require the agent to present costs, arguments, and total estimates to the user for confirmation before execution, preventing unauthorized resource consumption.
- [SAFE]: Data handling instructions emphasize sending the "minimum necessary data" and maintaining an audit trail, which reduces the risk of accidental data exposure.
Audit Metadata