on-device-ai
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- PROMPT_INJECTION (SAFE): The skill instructions include standard steering directives to ensure the agent uses provided reference files as the source of truth. There are no attempts to bypass safety filters or extract system prompts.
- DATA_EXFILTRATION (SAFE): No sensitive file paths, hardcoded credentials, or external network operations were detected.
- REMOTE_CODE_EXECUTION (SAFE): The skill does not perform any remote script downloads or dynamic code execution. It uses static regex patterns to validate user-provided code against best practices.
- INDIRECT_PROMPT_INJECTION (LOW): The skill is designed to analyze user-provided code/requests using validation rules in
references/validations.md. While this constitutes an ingestion point for untrusted data, the skill lacks capabilities (e.g., file writing, network access) that would allow for exploitation via this vector.
Audit Metadata