Episode 19 February 13, 2026

When guardrails break prod: GitHub “Too Many Requests” from legacy defenses, Kubernetes nodes/proxy GET RCE, HCP Vault resilience in an AWS regional outage, and PCI DSS scope creep

This week on Ship It Weekly, Brian hits four stories where the guardrails become the incident.

GitHub had “Too Many Requests” caused by legacy abuse protections that outlived their moment. Takeaway: controls need owners, visibility, and a retirement plan.

Kubernetes has a nasty edge case where nodes/proxy GET can turn into command execution via WebSocket behavior. If you’ve ever handed out “telemetry” RBAC broadly, go audit it.

HashiCorp shared how HCP Vault handled a real AWS regional disruption: control plane wobbled, Dedicated data planes kept serving. Control plane vs data plane separation paying off.

AWS expanded its PCI DSS compliance package with more services and the Asia Pacific (Taipei) region. Scope changes don’t break prod today, but they turn into evidence churn later if you don’t standardize proof.

Human story: “reasonable assurance” turning into busywork.

Links

GitHub: When protections outlive their purpose (legacy defenses + lifecycle)

https://github.blog/engineering/infrastructure/when-protections-outlive-their-purpose-a-lesson-on-managing-defense-systems-at-scale/

Kubernetes nodes/proxy GET → RCE (analysis)

https://grahamhelton.com/blog/nodes-proxy-rce

OpenFaaS guidance / mitigation notes

https://www.openfaas.com/blog/kubernetes-node-proxy-rce/

HCP Vault resilience during real AWS regional outages

https://www.hashicorp.com/blog/how-resilient-is-hcp-vault-during-real-aws-regional-outages

AWS: Fall 2025 PCI DSS compliance package update

https://aws.amazon.com/blogs/security/fall-2025-pci-dss-compliance-package-available-now/