email-marketing
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists exclusively of markdown instructions and a static HTML/CSS example. It does not include any scripts (Python, JavaScript, shell) or binaries that could execute commands on the host system.
- [INDIRECT_PROMPT_INJECTION]: The skill instructions specify reading an external
DESIGN.mdfile to determine typography and color tokens. While this represents a data ingestion surface, the skill's capabilities are limited to generating text/HTML artifacts, presenting a negligible security risk. - Ingestion points: Reads
DESIGN.mdas context for design tokens (referenced inSKILL.md). - Boundary markers: None explicitly defined for the
DESIGN.mdcontent. - Capability inventory: The skill is designed to emit an HTML artifact. It has no access to network tools, file-writing capabilities, or subprocess execution.
- Sanitization: No specific sanitization or escaping of the design tokens is mentioned, as they are used for visual styling in generated code.
Audit Metadata