ai-agent-design
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of technical documentation and instructional code samples intended for developer education.
- [PROMPT_INJECTION]: Phrases typically associated with instruction overrides are used exclusively within a Python code example demonstrating how to implement an 'InputGuardrail' to block such attacks.
- [SAFE]: Python snippets for memory management, orchestration, and tool use are pedagogical and do not perform unauthorized system actions or command execution.
- [SAFE]: No credentials, sensitive file paths, or network exfiltration patterns were identified within the guide.
Audit Metadata