code-implementer

Pass

Audited by Gen Agent Trust Hub on Feb 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is a set of instructional guidelines for software development and does not contain any executable malicious code or dangerous instructions.
  • [PROMPT_INJECTION]: The instructions define a specific role and boundaries for the AI without attempting to bypass safety protocols or override core instructions.
  • [CREDENTIALS_UNSAFE]: The skill explicitly instructs against hardcoding credentials and provides examples of secure configuration management using environment variables.
  • [REMOTE_CODE_EXECUTION]: No patterns of downloading or executing remote code were found. Examples use established libraries like 'jsonwebtoken' in a descriptive context.
  • [DATA_EXFILTRATION]: There are no commands or instructions that attempt to access sensitive files or exfiltrate data to external domains.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 26, 2026, 10:07 AM