prompt-engineering

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a documentation resource providing templates and architectural guidance for prompt engineering.
  • [PROMPT_INJECTION]: The skill identifies prompt injection as an anti-pattern and provides specific instructions on how to mitigate it, such as separating user input from system prompts.
  • [CREDENTIALS_UNSAFE]: The skill includes safety warnings against including PII or secrets in prompts to prevent data exposure via provider logs.
  • [COMMAND_EXECUTION]: The provided code snippets are illustrative examples of Node.js logic for LLM orchestration and CI testing; they do not contain malicious commands or unauthorized system access.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 10:28 AM
Security Audit — agent-trust-hub — prompt-engineering