gemini-api-dev

Pass

Audited by Gen Agent Trust Hub on May 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • Package Installation: The skill provides standard commands for installing the Google Gemini SDKs (google-genai, @google/genai, google.golang.org/genai). These are official packages from the vendor and are used as intended for development purposes.
  • External Documentation Access: The skill references official documentation hosted at ai.google.dev. This is a well-known, trusted domain for this vendor's documentation. The instructions provide a structured way for the agent to fetch up-to-date information through an llms.txt index or via MCP tools if available.
  • Instructional Overrides: The skill uses strong language (e.g., "Critical Rules", "Always Apply", "override your training data"). While this pattern is sometimes seen in prompt injections, here it is used to ensure the agent uses current model versions and SDKs instead of outdated training data, which is a common requirement for rapidly evolving APIs.
Audit Metadata
Risk Level
SAFE
Analyzed
May 20, 2026, 12:05 PM
Security Audit — agent-trust-hub — gemini-api-dev