deliverability-monitoring

Pass

Audited by Gen Agent Trust Hub on Jul 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the dig utility to query DNS records for SPF, DKIM, and DMARC auditing as part of the authentication audit process in SKILL.md.\n- [COMMAND_EXECUTION]: Includes a localized Python script (scripts/check-output.py) which the agent executes to verify that its output follows the required template structure.\n- [EXTERNAL_DOWNLOADS]: Recommends monitoring external reputation services and platforms such as Google Postmaster Tools, Microsoft SNDS, and industry-standard DNS-based blacklists (e.g., Spamhaus, Barracuda).\n- [PROMPT_INJECTION]: The skill documents processes for ingesting external data which represents a theoretical indirect prompt injection surface.\n
  • Ingestion points: DNS TXT records retrieved via dig (Phase 2) and DMARC aggregate XML reports processed in Phase 6.\n
  • Boundary markers: No specific delimiters or "ignore embedded instructions" markers are defined for the ingestion of this external data.\n
  • Capability inventory: Access to shell commands (dig) and local script execution (check-output.py).\n
  • Sanitization: The skill does not explicitly define sanitization routines for the external DNS or XML content before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 18, 2026, 08:48 PM
Security Audit — agent-trust-hub — deliverability-monitoring