14-email-marketing-global
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is composed strictly of Markdown documentation and instructional templates. No executable scripts, shell commands, or network-enabled tools are utilized.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where the agent reads a local context file (
.agents/product-marketing-context-global.md) to populate product and regional details. - Ingestion points: File access to
.agents/product-marketing-context-global.mdas described in Step 0 of the workflow in SKILL.md. - Boundary markers: Absent; the agent is instructed to read the file content directly as trusted context.
- Capability inventory: No tools, command execution, or executable capabilities are exposed or used by the skill files.
- Sanitization: None; the skill relies on the inherent safety of the ingested context for generating text-only strategy documents.
Audit Metadata