behavioral-modes
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill is entirely instructional and focused on optimizing AI interaction styles.
- [PROMPT_INJECTION]: The skill uses formatting keywords such as 'CRITICAL' and 'IMPORTANT' to define internal stylistic preferences (e.g., being concise during code implementation). These do not represent attempts to override system safety guidelines or bypass established security constraints.
- [DATA_EXFILTRATION]: There are no indicators of data exfiltration. The skill does not perform network requests and uses standard, restricted tools (Read, Glob, Grep) for local file navigation as defined in the platform configuration.
- [EXTERNAL_DOWNLOADS]: The skill does not reference external URLs, remote scripts, or third-party package dependencies.
Audit Metadata