sag
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill configures an open-source homebrew package ('steipete/tap/sag') for text-to-speech tasks. It does not introduce runtime remote code execution hooks, nor does it attempt data exfiltration. Credentials management recommendations leverage standard environment variables ('ELEVENLABS_API_KEY'). No adversarial patterns or prompt injection techniques were found.
Audit Metadata