eval-rubric-citation-quality
Pass
Audited by Gen Agent Trust Hub on Jun 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: This skill provides a set of criteria for scoring AI legal outputs on citation quality. It is a text-only rubric with no associated scripts, tools, or executable commands.
- [SAFE]: There are no patterns of prompt injection, obfuscation, or data exfiltration. References to external legal databases are for human or AI verification instructions and do not involve automated network operations.
Audit Metadata