reference-originality
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and referenced method document focus on qualitative analysis and evidence collection for comparing design and expression. The methodology is well-documented and restricts itself to informational assessments rather than making legal claims.
- [EXTERNAL_DOWNLOADS]: The skill includes an informational link to a GitHub repository (MengTo/Skills) to cite the method's lineage. This is a static reference for attribution and does not involve automated code execution or package installation.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external subjects and references, which constitutes a data ingestion surface in
references/reference-originality-method.md. While instructions could be hidden in analyzed documents, the skill lacks dangerous capabilities (network access, file writing, or command execution), and the structured analysis process helps ensure the agent remains focused on the comparison task.
Audit Metadata