gemini-api-dev

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill includes directives that claim the agent's internal knowledge is outdated and mandates the use of specific model versions while declaring current real-world versions as deprecated. This is a technique used to override the agent's baseline training data.
  • [DATA_EXFILTRATION]: (Safe) Fetches API discovery specifications and documentation from official Google domains including googleapis.com and ai.google.dev for legitimate integration purposes.
  • [EXTERNAL_DOWNLOADS]: (Safe) The skill recommends the installation of official Gemini SDKs such as the google-genai Python package and the @google/genai Node.js package from standard registries.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 11:06 PM