emails
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is composed strictly of Markdown instructions and reference files. No executable scripts, automated tasks, or binaries are included in the skill files.
- [SAFE]: External tool references (Mailchimp, Resend, etc.) are limited to informational guides and standard tool integration registries, with no automated command execution.
- [INDIRECT_PROMPT_INJECTION]: The skill intentionally reads from local context files like
.agents/product-marketing.mdto tailor its output to the user's business. - Ingestion points:
.agents/product-marketing.md,.claude/product-marketing.md, and user-provided inputs in SKILL.md. - Boundary markers: Absent. The skill does not define specific delimiters to isolate external context from system instructions.
- Capability inventory: Text generation (email copy) and referencing external implementation guides in SKILL.md. No file-writing or network capabilities are active in the skill instructions.
- Sanitization: Absent. The skill relies on the underlying LLM's default safety guardrails when processing ingested product marketing data.
Audit Metadata