← ClaudeAtlas

cloud-native-security-best-practiceslisted

Design and review cloud-native security across workloads, images, identities, networks, secrets, supply chains, runtime controls, and recovery.
sandbaseai/workbuddy-skill · ★ 9 · AI & Automation · score 80
Install: claude install-skill sandbaseai/workbuddy-skill
# Cloud-Native Security Best Practices Use this skill when reviewing or designing security for containerized, orchestrated, serverless, or otherwise distributed cloud workloads. Start from assets, trust boundaries, and likely failure or abuse paths rather than from a vendor checklist. The skill provides a review method; it does not claim access to a cluster, cloud account, image registry, or policy engine unless an authorized read-only integration actually supplied evidence. ## Define the security contract Record the application, environment, cluster or account boundary, workload owners, data sensitivity, compliance obligations, deployment model, and review commit. Map: - identities for users, workloads, nodes, controllers, CI, and operators; - image, dependency, build, registry, admission, and deployment supply chain; - network ingress, egress, service-to-service, control-plane, and data-plane boundaries; - secrets, keys, certificates, config, persistent data, and backup paths; - runtime capabilities, host access, namespaces, tenancy, and blast radius; - detection, response, recovery, and evidence-retention ownership. Separate observed platform facts from assumptions. A cloud provider default, manifest example, or dashboard green state is not proof that the deployed workload has the intended control. ## Apply layered controls Review controls in this order, keeping each control tied to an abuse path and an owner: 1. **Identity and authorization:** use short-lived, a