routing-strategies

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • SAFE (SAFE): No malicious behaviors or security risks were identified across the 10 threat categories.
  • Data Exposure & Exfiltration: The skill uses environment variables for broker URLs (CELERY_BROKER_URL) and provides a script to specifically scan for and alert on hardcoded credentials in configuration files.
  • External Downloads & RCE: No external code is downloaded or executed at runtime. Network operations in the test scripts are limited to connectivity checks against localhost and standard broker ports.
  • Privilege Escalation & Persistence: The skill does not attempt to escalate privileges or establish persistence. The use of sudo in documentation is restricted to administrative commands for RabbitMQ management (rabbitmqctl).
  • Logic Analysis: The routing logic provided in the templates uses safe mapping via dictionaries and string prefix checks, which effectively sanitizes input task names and arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:50 PM