cmo-marketing

Pass

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill serves as a high-level controller that ingests and processes untrusted external data (such as competitor landscapes) to drive strategy and direct downstream agents.\n- Ingestion points: The Inputs section in SKILL.md describes the ingestion of competitor landscape briefs, market positioning data, and business goals which are often derived from external sources.\n- Boundary markers: The skill's instructions do not define clear delimiters or use explicit prompt techniques to isolate untrusted research data from the core mission instructions, creating a surface where embedded instructions in research data could potentially override agent behavior.\n- Capability inventory: As a top-level orchestrator, this skill has the authority to commission a wide array of functional sub-agents, including content-marketing, demand-gen, paid-ads-manager, and customer-intel.\n- Sanitization: The skill lacks any explicit logic for sanitizing or validating the content of the business intelligence documents before they are used to generate quarterly strategy briefs.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 9, 2026, 05:42 AM