agent-guardrails-test
Pass
Audited by Gen Agent Trust Hub on Apr 15, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local bash script at .claude/hooks/stop-guardrails.sh to verify that security guardrails correctly handle specific test phrases.
- [COMMAND_EXECUTION]: The skill uses cat and jq to read and parse the local configuration file .claude/settings.local.json to ensure the hook is correctly registered before running tests.
Audit Metadata