Cut the time it takes to understand why production is broken
Kunobi puts pod state, events, logs, and what changed last in one window, even when your VPN to the bastion is half-broken. Keyboard-first. Works offline.
$ kunobi logs --multi payments-api
▌pod-a 14:23:11 ERROR connection refused
▌pod-b 14:23:11 ERROR connection refused
▌pod-c 14:23:12 ERROR connection refused
$ kunobi events --since 5m
WARNING Unhealthy pod/payments-api Liveness probe failed (3×)
NORMAL Killing pod/payments-api Container restart
NORMAL Scheduled pod/payments-api-...
# what changed and when, in one windowOutcomes, not feature lists
Pod state, events, and logs in one drilldown
Stop alt-tabbing. Open a pod and see its current phase, restart counter, last events, and live logs, all in the same window. The story of the failure on one screen.
Multi-stream logs without picking a pod
Aggregate logs across every pod of a deployment, every container of a pod, every revision of a rollout. Filter and search in real time. No more `kubectl logs -f` lottery.
Rolling updates and rollbacks made visible
See the ReplicaSet handover happen as it happens, old pods scaling down, new pods coming up. Roll back to any previous revision with Shift+B. No more guessing which revision was good.
Works when your network doesn’t
Recent cluster state stays queryable on your laptop when the VPN to the bastion flakes. Keyboard-first, every list reachable without the mouse. Stay in the home row, stay in the fight.
Get it on your laptop before the next incident
No login. No signup. No agents. Free for individual SRE use. Point it at the kubeconfig you already have and start working, even from a coffee shop.
Lower MTTR. Quieter on-calls
Cut the alt-tab tax during incidents. Make incident response repeatable and onboarding-friendly. The cluster state your on-call needs is already on their laptop, even when nothing else is reachable.