Skip to content
Notifications
Clear all

Checkmarx rollout in a K8s CI/CD pipeline - integration pitfalls

1 Posts
1 Users
0 Reactions
3 Views
(@cloud_security_sera)
Estimable Member
Joined: 1 month ago
Posts: 134
Topic starter   [#2317]

Just finished rolling out Checkmarx SAST into our Kubernetes CI/CD. The default "integrations" are a security hazard.

The main pitfalls aren't in the scanner itself, but in the permissions and data flow everyone glosses over.

* **IAM/Service Account Over-provisioning:** The Helm chart/service connector often needs cluster-admin or wildcard permissions to "make it easy." Never.
```yaml
# Bad: Default suggestion
rules:
- apiGroups: ["*"]
resources: ["*"]
verbs: ["*"]
# Better: Minimal RBAC
- apiGroups: [""]
resources: ["pods", "pods/log"]
verbs: ["get", "list", "create"]
```
* **Secret sprawl in CI variables:** Embedding the Checkmarx API token as a plain environment variable in your pipeline. Any compromised build container can exfiltrate it.
* **Result handling creates shadow data lakes:** Findings dumped to a pipeline artifact or a public S3 bucket without encryption/logging. Now you have an unaccounted PII/data source.

How are teams handling the service account rotation and ensuring scan results are only accessible to the security/SOC tooling? The vendor docs are silent on the operational security controls.


Least privilege is not a suggestion.


   
Quote