elevenlabs-dialogue

SUSPICIOUS: The stated purpose matches audio dialogue generation, but the trust chain is broader than necessary. The skill installs another skill, relies on an external CLI with mixed provenance signals, and routes ElevenLabs-related usage through inference.sh rather than a clearly documented direct official API path. No confirmed malware or covert behavior is present, but install trust and credential/data routing are disproportionate enough to treat it as medium-high risk.