backend-implement

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill mandates high security standards for all implemented code, including explicit requirements for input validation, query parameterization to prevent injection, and protection of sensitive data like PII.
  • [SAFE]: Critical and potentially destructive operations, such as database migrations, authentication logic, and payment processing, require mandatory human intervention and approval.
  • [SAFE]: Clear safety constraints are established to prevent the execution of destructive commands (e.g., rm -rf, DROP TABLE) and the handling of production secrets by the agent.
  • [SAFE]: Context loading practices include the use of specific delimiters to separate file contents, mitigating risks of context confusion or indirect instruction interference when processing multiple files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 09:27 AM
Security Audit — agent-trust-hub — backend-implement