goal-loop-builder

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local Python script, scripts/lint_goal_loop.py, to validate generated artifacts. Analysis of the script shows it uses only standard libraries for static text analysis and contains no network operations, subprocess spawning, or unsafe dynamic execution.
  • [SAFE]: Security-conscious instructions: The skill's workflow and templates mandate the inclusion of 'Boundaries' (allowed/forbidden paths) and 'Pause' conditions for high-risk scenarios involving credentials, production data, destructive actions, or financial judgments.
  • [SAFE]: No malicious patterns: The skill does not contain any prompt injection attempts, obfuscated content, hardcoded credentials, or unauthorized external connections.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 11:42 AM
Security Audit — agent-trust-hub — goal-loop-builder