Configure the AI Policy Engine
Set risk thresholds and approval requirements for every type of autonomous action.
Configuring the AI Policy Engine
- Go to Settings → AI Autonomy → Policy Engine.
- Each action type (e.g.
diag.run-script,ticket.add-comment) has a risk level:- Low — executes automatically
- Medium — executes after a delay window; approval cancels it
- High — requires explicit approval before execution
- Click an action type, set the risk level, and save.
Approval workflow
High-risk actions create a pending approval request. Approvers receive an email with a one-click approve/reject link. If no action is taken within the timeout (default 30 min), the action is rejected.
All decisions are logged under Settings → AI Autonomy → Audit Log.
Where to go next
- Understand the whole module → How AI Autonomy Works
- Next in this module → Use AI Ticket Triage
- Diagnostic Playbooks
- Run a Diagnostic Playbook
- See where this fits in the bigger picture → Start Here: What Plenix Is and How It Fits Together
Was this article helpful?