If you need the exact list of policy identifiers used internally (e.g., HATE_SPEECH, SELF_HARM, ILLICIT_ACTIVITY, etc.), let me know and I can provide the short code names as well.

{
"guardrails": [
{"name":"self_harm","status":"passed"},
{"name":"hate_speech","status":"blocked","action":"safe_completion"},
{"name":"privacy","status":"passed"}
],
"safe_completion_message":"I’m sorry, but I can’t help with that."
}