deliverability-qa
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes local Python scripts located within the plugin root directory (
resend.pyanddoh.py) to retrieve domain authentication status and perform DNS-over-HTTPS lookups. These commands are used solely for the skill's primary purpose of deliverability auditing. - [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it is designed to process untrusted data from external sources.
- Ingestion points: The skill ingests user-provided DMARC aggregate reports, DNS exports, ESP deliverability reports, and creative campaign HTML (
SKILL.md). - Boundary markers: The instructions explicitly warn the agent to treat every exported file and report as untrusted input and to treat text inside these reports as evidence rather than commands (
SKILL.md). - Capability inventory: The skill utilizes shell execution for internal connector scripts and has the ability to write analysis results to the
memory/directory. - Sanitization: The skill incorporates a safety directive referencing
SECURITY.mdto ensure that instructions embedded within analyzed data do not override agent behavior.
Audit Metadata