email-audit
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and interprets text from public DNS records (SPF, DMARC, DKIM, BIMI), which represents a surface for untrusted data from external sources. An adversary could potentially embed prompt injection payloads within these records to attempt to influence the agent's analysis or behavior.
- Ingestion points: The skill ingests data via the
email_tools_audittool as described inSKILL.md, which resolves live records from the internet. - Boundary markers: No specific delimiters or instructions to ignore embedded instructions are present in the instructions to protect the agent's context during the report interpretation phase.
- Capability inventory: The skill is designed to use the
email_tools_audittool to perform DNS lookups on arbitrary domains provided by the user. - Sanitization: The instructions do not include any steps for sanitizing, escaping, or validating the content of the resolved DNS strings before they are analyzed by the agent.
Audit Metadata