[VC-35565] Automate the e2e script to run in CI

[mladen-rusev-cyberark]

1. Add the test.sh script to be run by the CI
2. Add all dependencies required to execute the script in extra_tools.mk
3. Created a GCP account (via SNOW request) and setup inside it:

• A service account with manage GKE permissions, Write Image Repository Permissions and the relevant APIs enabled.
• A JSON auth key for this account was set as a secret to the CI runner in GCP_SA_KEY
• A repository to push the images which the script builds

4. A valid API key for https://glow-in-the-dark.venafi.cloud/ was set as a CI secret in VEN_API_KEY_PULL
5. A valid VEN_ZONE was configured - mladen-test\Default. If this is not valid test would fail because the certificate request would be ignored.
6. Created two labels: test-e2e and keep-e2e-cluster - one to trigger the e2e test and the other to keep the cluster alive in GKE for debugging.
7. Reduced log level for cleaner logs.

time=2025-09-15T15:56:32.291Z level=DEBUG+3 msg="Request is not applicable for any policy so ignoring" logger=controller-manager/events type=Normal object="{Kind:CertificateRequest Namespace:team-1 Name:app-0-1 UID:12206e29-27b6-4ea4-8171-5c335fc61537 APIVersion:cert-manager.io/v1 ResourceVersion:1757948283819711024 FieldPath:}" reason=Unprocessed

Closes https://venafi.atlassian.net/browse/VC-35565

[wallrj-cyberark]

A few suggestions for things to be addressed now or in future PRs.

[wallrj-cyberark]

Thanks @mladen-rusev-cyberark

Lets merge this and try it out and improve it in future PRs.

For example, I think we should probable add some labels to the GKE cluster so that the security team using Wiz.io can trace this back to our team and understand the purpose of the cluster.

Also have no idea how we are going to turn on coverage instrumentation in the agent binary and then copy the coverage report back from the agent running in this remote cluster.

[wallrj-cyberark]

So does makefile-modules automatically add _bin/tools to the PATH?

[mladen-rusev-cyberark]

I think so