teaching-evaluating-reason-about

Pass

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides domain-specific instructions for materials science reasoning. No malicious commands, obfuscated code, or unauthorized data access patterns were found.
  • [PROMPT_INJECTION]: The content does not attempt to override agent safety protocols or system prompts. Instructional keywords like "IMPORTANT" are not present in a malicious context.
  • [DATA_EXFILTRATION]: There are no indicators of unauthorized data collection or network exfiltration to untrusted domains. The use of polymer databases and user data is described within a legitimate scientific workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 19, 2026, 04:04 PM
Security Audit — agent-trust-hub — teaching-evaluating-reason-about