validating-business-logic-against-code
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill primarily uses standard, non-destructive commands like
catandgrepto analyze local project files. - [SAFE]: No external network operations (exfiltration) or remote code downloads (RCE) were detected in the instructions or examples.
- [SAFE]: The skill does not request or use high-privilege commands like
sudoor modify sensitive system configurations. - [SAFE]: The use of
allowed-tools(Read, Grep, Glob) is restrictive and well-aligned with the task of static code and document analysis. - [SAFE]: While the skill processes external data (business logic documents and code), it does not execute this data as code, which significantly mitigates indirect prompt injection risks.
Audit Metadata