guv3-prompt-patterns
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill establishes secure prompt engineering patterns, including explicit grounding instructions to prevent hallucinations ('Solo responde con datos explicitos', 'NUNCA inventes datos') and clear role separation definitions for assistants.\n- [NO_CODE]: The skill consists entirely of documentation and illustrative code snippets for reference. It does not contain standalone executable logic, installation scripts, or automated routines that could be exploited.
Audit Metadata