prompt-engineering
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains no executable code or scripts. It consists of Markdown documentation providing guidance and best practices for prompt engineering.
- [SAFE]: Data exposure risks are addressed through explicit instructions to avoid storing sensitive information like passwords or API keys in agent memory systems.
- [SAFE]: The skill provides defense-in-depth guidance, including patterns for prompt injection defense, such as distinguishing between system instructions and untrusted data from memories or web searches.
- [SAFE]: No obfuscation, remote code execution patterns, or unauthorized persistence mechanisms were detected.
Audit Metadata