loss-aversion-designer
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of natural language instructions for marketing strategy and copy planning. It emphasizes ethical guardrails, avoiding fear-mongering, and preserving user autonomy.
- [NO_CODE]: There are no scripts, commands, or executable segments within the skill. It operates purely as a set of logical and psychological frameworks for the agent to follow.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes user-provided inputs (e.g., product offers or messages), it lacks the technical capabilities (such as network access, file system modification, or command execution) to facilitate a meaningful attack. It includes explicit checks for output quality and ethical boundaries.
Audit Metadata