skills/alekspetrov/navigator/nav-loop/Gen Agent Trust Hub

nav-loop

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The unit tests use subprocess.run to execute the skill's own Python scripts for integration testing. This is a standard and safe practice for testing CLI interfaces with controlled inputs.
  • [DATA_EXPOSURE] (SAFE): No patterns of sensitive data access (e.g., credentials, SSH keys) or network exfiltration were detected.
  • [REMOTE_CODE_EXECUTION] (SAFE): The code exclusively uses the Python standard library and does not download or execute remote code or untrusted third-party dependencies.
  • [PROMPT_INJECTION] (SAFE): No instructions designed to override agent behavior or bypass safety guardrails were found. The skill focuses on objective state tracking and reporting.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:32 PM