tw/commands: add a new functionality or tool to expect a particular Pods to spin up in a certain period

Some Kubernetes Operators spins up Pods after we apply Custom Resource. But it usually took a while during reconciliation and get healthy status. So the Pods don’t exist yet for a certain period.

Example:

wait_for_existence -n foo 'app.kubernetes.io/name=bar' --timeout=5m

This command would execute the followings in order:

1. Set timeout to N seconds and interval to M seconds
2. Loop until timeout is reached
3. Brute-force get pod status with kubectl get pods with selection filter
4. If no pods exist yet, print waiting message
5. If pods exist (on matching), check for Running status with kubectl wait
6. If a match is found, and healthy, print success message and exit
7. Otherwise, print waiting message
8. Sleep for the interval and increment elapsed time
9. timeout is reached, print failure message and exit

cc @joshrwolf

Fun fact: This is the FIRST ISSUE!

Start by inspecting the tw/commands entry points and existing command patterns. Implement the described polling flow for a namespace, label selector, interval, and timeout, then verify that it waits for matching Pods to become Running and reports success or failure as specified.