fede-voice
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data such as message threads and notes (SKILL.md). While there are no boundary markers, the skill lacks any functional capabilities such as tool usage, shell execution, or network operations that could be exploited via injection. The instructions include manual sanitization by requiring the agent to verify all facts, names, and dates against original sources (SKILL.md).
- [SAFE]: No other security concerns, such as credential exposure, malicious dependencies, or obfuscation, were detected.
Audit Metadata