codeguard-arkts
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill explicitly defines a 'read-only' posture and prohibits the collection, storage, or transmission of user code, tokens, or credentials.
- [SAFE]: No obfuscation, remote code execution, or unauthorized network operations were detected. The skill specifically forbids silent or background installation of dependencies.
- [SAFE]: The workflow requires human intervention, including diff reviews and explicit confirmation before any code modification, which protects against unintended side effects during automated fixing attempts.
- [SAFE]: The skill correctly identifies its own lack of executable linting tools (status: planned) and instructs the agent not to fabricate or simulate check results, ensuring evidence-based reporting.
Audit Metadata