model-serving

Warn

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: MEDIUMREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The provided code uses 'joblib.load()' to deserialize model artifacts. Since joblib is based on the pickle protocol, it can execute arbitrary code embedded in a malicious file. The skill documentation explicitly flags this as a pitfall to avoid.
  • [COMMAND_EXECUTION]: The documentation includes instructions to execute the 'uvicorn' command, which launches a web server and binds it to all network interfaces.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 14, 2026, 04:55 AM