graphrag-evaluation
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists exclusively of instructional markdown and a JSON-based rubric. All resources are locally hosted within the skill folder. No shell commands, network exfiltration, or obfuscated payloads are present.
- [SAFE]: The skill provides a methodology for evaluating RAG outputs which are externally sourced. However, because the skill does not grant the agent any high-privilege capabilities (like file writing or network access), no actionable indirect injection vulnerability is created.
Audit Metadata