tax-withholding-review

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were identified in the skill definition.
  • [PROMPT_INJECTION]: The instructions do not attempt to bypass safety guidelines, assume unrestricted roles, or manipulate the agent's core instructions.
  • [DATA_EXFILTRATION]: The skill processes payroll data locally as part of its intended finance auditing function. No network requests, external data transfers, or hardcoded credentials were found.
  • [REMOTE_CODE_EXECUTION]: The file does not include any external package dependencies, remote script downloads, or dynamic code execution logic.
  • [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or persistent mechanisms were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 02:26 AM
Security Audit — agent-trust-hub — tax-withholding-review