justinian-legal-essay-grader

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or dangerous capabilities were detected. The skill consists entirely of instructional text and metadata.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data (student essays and fact patterns). While there are no explicit boundary markers or instructions to ignore embedded commands within the input, the skill possesses no capabilities (such as shell access, network requests, or file writing) that could be exploited via injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 02:24 PM
Security Audit — agent-trust-hub — justinian-legal-essay-grader