voice-clone-lab
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes user-provided consent metadata and audio samples, creating an indirect prompt injection surface.\n
- Ingestion points: User-supplied consent text and audio sample paths extracted from requests in
SKILL.md.\n - Boundary markers: No specific delimiters or markers are defined to isolate user-provided data from the agent's instructions.\n
- Capability inventory: The skill utilizes the
voice_cloneandaudio_provider_capabilitiestools and declaresnetwork-readandfilesystem-writecapabilities in its metadata.\n - Sanitization: The instructions explicitly require the agent to validate the presence and clarity of consent metadata before tool invocation, acting as a manual validation step for external inputs.
Audit Metadata