creativeclaw-clone-voice
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill mandates a robust consent workflow. It requires the agent to obtain explicit verbal agreement to a detailed legal and privacy statement before any voice cloning or recording operations are initiated.
- [SAFE]: Data protection instructions are prominently featured. The skill explicitly forbids the agent from sending private data, local system paths, or information about other plugins to the external tools.
- [SAFE]: The skill utilizes a vendor-specific MCP server (creative-claw) for its operations. All tool calls and data uploads (e.g.,
import_media,clone_voice) are conducted through this authenticated and scoped infrastructure. - [SAFE]: External service integrations (Cartesia, ElevenLabs, Fish Audio, HeyGen, MiniMax, xAI) are limited to well-known, industry-standard AI providers directly related to the skill's purpose. The links provided point to official documentation and privacy policies on these providers' domains.
- [SAFE]: The skill instructions emphasize the use of native niqqud (vowel points) for Hebrew speech synthesis, which is a legitimate technical requirement for TTS engines and does not constitute a homoglyph or obfuscation attack.
Audit Metadata