content-scoring

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, hidden instructions, or dangerous command executions were found. The skill operates entirely as a text-based evaluation framework.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied content to generate scores. While this represents a data ingestion surface, the skill does not have access to any external tools, network resources, or sensitive file paths that could be exploited via injected instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 05:16 AM
Security Audit — agent-trust-hub — content-scoring