writing-north-star-metrics

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION] (SAFE): No override markers, role-play injections, or system prompt extraction attempts were detected. The instructions are purely functional for the task of metric definition.
  • [DATA_EXFILTRATION] (SAFE): No hardcoded credentials, sensitive file paths, or network operations (curl, wget, fetch) are present. The skill does not interact with external APIs or the local filesystem.
  • [REMOTE_CODE_EXECUTION] (SAFE): There are no external dependencies, package installations, or remote script download patterns.
  • [COMMAND_EXECUTION] (SAFE): No shell commands or system-level calls are used. The skill operates within the context of text generation and template filling.
  • [NO_CODE] (SAFE): The skill is composed strictly of Markdown documentation and structural metadata, presenting zero attack surface for runtime execution threats.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:27 PM