hayek-perspective
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and provides a personality/analytical framework for the AI. No malicious logic was found.
- [PROMPT_INJECTION]: The instructions do not attempt to bypass safety filters, override system prompts, or extract internal instructions. The persona is defined clearly within the intended bounds.
- [DATA_EXFILTRATION]: No commands for accessing sensitive files (e.g., .ssh, .aws) or making unauthorized network requests were found.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code execution, script downloads, or package installations. The 'npx' command mentioned in the README is for platform-level skill installation, not runtime execution.
- [OBFUSCATION]: No hidden content, Base64 encoding, or zero-width characters were detected in any of the analyzed files.
- [DYNAMIC_CONTEXT_INJECTION]: The skill does not use any shell execution syntax (e.g., backticks with exclamation marks) in its configuration files.
Audit Metadata