goal
Pass
Audited by Gen Agent Trust Hub on Jun 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill enhances agent security by enforcing the creation of 'Goal Contracts' that define explicit write boundaries, constraints, and pause conditions (e.g., pausing for credentials or destructive actions) before any work begins.
- [COMMAND_EXECUTION]: The skill incorporates a local Python script (
check_goal_contract.py) used to lint generated contracts. This script was analyzed and verified as a safe text validator that uses only the Python standard library to ensure contract quality. - [SAFE]: No malicious patterns, such as prompt injection, data exfiltration, or obfuscation, were detected within the skill files or metadata.
Audit Metadata