building-ai-agents

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is comprised entirely of markdown reference files and documentation. It does not contain any executable scripts, binaries, or automated tasks that could pose a security risk.
  • [PROMPT_INJECTION]: Static analysis detected strings associated with prompt injection (e.g., "ignore previous instructions") in references/ADK-SECURITY-AND-GOVERNANCE.md. However, these are used exclusively as educational examples in a security guide to demonstrate how to build robust system prompts that resist injection attacks. They are not active injection attempts.
  • [EXTERNAL_DOWNLOADS]: The guides mention several well-known libraries and frameworks (e.g., google-adk, langchain, litellm) and provide standard installation commands. These are references to trusted, official development tools and do not involve suspicious remote code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 11:13 PM