genmedia-voice-director
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill acts as an instructional template for controlling Gemini TTS via the
gemini_audio_ttstool. Its operations are limited to prompt construction for audio generation, which is a benign and intended use case. All referenced tools and frameworks are standard for multimodal AI development. - [EXTERNAL_DOWNLOADS]: The skill links to official Google Cloud and Gemini API documentation in the frontmatter (
ai.google.dev,docs.cloud.google.com). These references are used for developer guidance and originate from a trusted organization. - [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to populate a voiceover transcript. While this creates an injection surface where user-provided text is passed to a downstream TTS model, the risk is mitigated by the structure of the prompt (using a
#### TRANSCRIPTheader) and the fact that the output is limited to generated audio. No evidence of data exfiltration or command execution capabilities was found in the associated scripts.
Audit Metadata