The Agent Skills Directory

[Data Sanitization]: The skill includes a dedicated JQ script (sanitize-readiness-input.jq) designed to redact sensitive information from Kubernetes cluster data. This proactive measure ensures that when the agent reads workload configurations, fields containing potentially sensitive data—such as environment variable values and secret references—are removed before being processed.
[Secure Resource Management]: The skill emphasizes the use of Azure Workload Identity and Azure Key Vault for managing credentials. It explicitly instructs the agent not to request or output secrets, tokens, or subscription identifiers, adhering to the principle of least privilege and minimizing the risk of credential exposure.
[Trusted External References]: The provided templates and instructions reference well-known and trusted technology providers. It uses official container images from Microsoft Container Registry and Google Container Registry, and downloads standardized definitions from the official Kubernetes-SIGs repository.
[Workload Integrity]: The deployment workflows include validation checks against AKS Deployment Safeguards (DS001–DS013). These checks enforce security configurations like resource limits, liveness/readiness probes, and non-root execution for all containerized applications.

azure-kubernetes