verbose-error-messages-anti-pattern

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOW
Full Analysis
  • [Metadata Poisoning] (SAFE): Metadata accurately reflects the content and purpose of the skill as a security educational resource.
  • [Unverifiable Dependencies & Remote Code Execution] (SAFE): No external code is downloaded or executed; the provided examples use standard libraries like logging and traceback.
  • [Prompt Injection] (SAFE): No instructions designed to override agent safety protocols or bypass constraints were detected.
  • [Indirect Prompt Injection] (SAFE): The skill contains static documentation and does not ingest or process untrusted external data in an automated or exploitable way.
  • [Data Exposure & Exfiltration] (SAFE): No sensitive file access or unauthorized network operations are present.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 12:57 PM