gemini-api-dev
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill includes instructions designed to override the agent's internal knowledge base by falsely claiming that existing Gemini models (1.5, 2.0) are legacy and deprecated. It asserts that the agent's knowledge is outdated and directs it to use hallucinated versions like 'gemini-3-pro-preview'.
- [EXTERNAL_DOWNLOADS]: The skill instructs the user to install the 'google-genai' Python package and '@google/genai' Node.js package. While these are legitimate SDKs from a trusted organization (Google), the surrounding instructions provide misleading context regarding the deprecation of older SDK versions.
- [PROMPT_INJECTION]: The skill points the agent to ingest external documentation from 'ai.google.dev/gemini-api/docs/llms.txt'. This represents an indirect prompt injection surface where the agent is encouraged to fetch and process external data, although the specified domain is recognized as a well-known service from a trusted provider.
Audit Metadata