Agentic Operations ● continuous

The AI agents that audit the network non-stop and generate reports. They run on a schedule and react to config-change events — "continuous" = frequent sweeps + event triggers, not magic real-time.

Run history

All runs →

Autonomy level

Policy →
L1
L2
L3
L4
L5
Currently L3 · conditional: agents plan & execute audits autonomously, but hold risky actions (remediation, high-severity) for human approval. Low-risk fixes run at L4.

Approval queue

Remediation →
Revert PE-JKT-03 to golden v14conf 0.97 · exceeds L4 threshold but high-severity → human sign-off
Firmware rollback CE-MDN-22service-affecting · awaiting approval