21-deliverability-sender-operations
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override safety filters or agent behavior were found. The instructions are focused on operational safety and compliance.
- [DATA_EXFILTRATION]: No sensitive file paths, hardcoded credentials, or unauthorized network operations were detected. References to DNS records and email provider APIs are within the scope of the skill's purpose.
- [REMOTE_CODE_EXECUTION]: No script downloads, remote execution patterns, or dynamic code generation were found.
- [COMMAND_EXECUTION]: The skill does not instruct the agent to execute arbitrary system commands, acquire elevated privileges, or modify system persistence mechanisms.
- [EXTERNAL_DOWNLOADS]: No external software dependencies or script downloads were identified. Citations to documentation from well-known services (Google, Yahoo, Microsoft, AWS) are provided for reference and are safe.
- [CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or passwords were found in the instructions or metadata.
Audit Metadata