ecommerce-copywriter
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security violations were detected. The skill operates as a pure prompt-based assistant without external tool access or script execution capabilities.
- [INDIRECT_PROMPT_INJECTION]: The skill contains an attack surface for indirect prompt injection as it processes untrusted user data to generate copy.
- Ingestion points: The skill ingests user-supplied product names, brands, price points, and selling points to formulate marketing text (defined in SKILL.md).
- Boundary markers: Absent; there are no specific delimiters or instructions to ignore embedded commands in the user input.
- Capability inventory: No capabilities detected. The skill does not use subprocesses, file writes, network operations, or external tools.
- Sanitization: No validation or sanitization of user-provided strings is performed.
Audit Metadata