gemini-live-api-dev
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides documentation and code examples for the Gemini Live API. All identified resources, including Python and Node.js packages, trace back to the official vendor (Google) or well-known and reputable service providers in the developer ecosystem.
- [SAFE]: External links provided in the documentation point to official Google Cloud/AI domains or established technology companies such as LiveKit, Daily (Pipecat), Software Mansion (Fishjam), Stream, and Voximplant.
- [SAFE]: Code examples use placeholders for sensitive information (e.g.,
YOUR_API_KEY) and follow standard security practices by recommending ephemeral tokens for client-side authentication. - [INDIRECT_PROMPT_INJECTION]: The skill documents an interface for processing untrusted real-time user input (audio, video, text). This represents a standard attack surface for AI applications. The skill includes instructions for system prompts and function calling, which are standard API features. Severity is assessed as safe as the skill is an informational resource and does not implement unsafe handling of this data itself.
Audit Metadata