ideogram-reference-architecture

Fail

Audited by Gen Agent Trust Hub on Feb 18, 2026

Risk Level: CRITICAL
Full Analysis
  • Prompt Injection (SAFE): No instructions attempting to override agent behavior or bypass safety guidelines were found.
  • Data Exposure & Exfiltration (SAFE): No hardcoded credentials or sensitive file access patterns detected. Configuration management uses standard environment variables rather than hardcoded secrets.
  • Unverifiable Dependencies & Remote Code Execution (SAFE): The provided shell scripts only create local directories and empty files. No external packages or remote scripts are downloaded or executed.
  • Malicious URLs (SAFE): The reported automated alert for 'this.ca' is a false positive triggered by the TypeScript code snippet this.cache. The scanner incorrectly interpreted a code property access as a malicious domain.
  • Dynamic Execution (SAFE): While the example code uses dynamic require for configuration files, it is restricted to local JSON files based on the environment name and is provided as a user template rather than executable skill code.
Recommendations
  • Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Feb 18, 2026, 09:00 PM