deliverability-monitoring
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze external data from email bounce exports and DMARC aggregate reports in XML format. This creates an attack surface where instructions hidden in the processed data could influence the agent's behavior. \n
- Ingestion points: Phase 3 (Bounce Rate Analysis) and Phase 6 (DMARC Aggregate Report Analysis) in SKILL.md. \n
- Boundary markers: No specific delimiters or safety instructions are defined for the agent to use when parsing these external files. \n
- Capability inventory: The skill uses the dig command and directs the agent to parse and categorize structured data from multiple external sources. \n
- Sanitization: The instructions do not include steps for sanitizing or validating the contents of the imported reports. \n- [COMMAND_EXECUTION]: The skill utilizes the dig and grep command-line utilities to perform DNS lookups for SPF, DKIM, and DMARC verification. These are standard diagnostic operations for deliverability monitoring.
Audit Metadata