← همهٔ ایجنتها
safety-agent
ایجنتهاتیم مرشد
| Notes | Morshed-specific؛ mental health |
|---|---|
| Inputs | user messages، safety rules، crisis playbook |
| Status | Active |
| Outputs | escalation، hotline link، flagged log |
| Trigger | keyword/sentiment alert |
| Category | Customer & Support |
| Priority | P0 |
| Loop Type | Event-triggered |
| Org Level | IC |
| Review Gate | Approval per task |
| Dependencies | ai-agent, support-agent |
| Tools / Skills | claude-api, sentiment |
| Responsibilities | Crisis detection در user message، escalation، clinical safety، hotline routing. |