commerce-product-naming
Pass
Audited by Gen Agent Trust Hub on Aug 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: Analysis of the skill instructions and metadata confirms the absence of malicious code, obfuscation, or persistence mechanisms. The behavior is consistent with the stated purpose of e-commerce name optimization.
- [PROMPT_INJECTION]: The skill is designed to ingest and process user-provided external files and keyword research data, creating a surface for indirect prompt injection.
- Ingestion points: The skill accepts user-attached research files (Excel/CSV) and text-based USP descriptions to guide name generation.
- Boundary markers: Output is requested in a structured JSON format, which helps define data boundaries, but there are no explicit instructions to ignore instructions potentially embedded within user data.
- Capability inventory: Capabilities are limited to text analysis and generation; the skill does not have access to sensitive local files or arbitrary shell execution.
- Sanitization: The workflow includes mandatory chaining with ai-slop-reviewer and korean-humanize tools, which act as filters for anomalous or low-quality generated content.
Audit Metadata