What is ChangeGuard?
ChangeGuard is a Kubernetes security platform that gives platform engineering teams complete visibility into the safety, compliance, and runtime health of every cluster. At its core is the Change Safety Confidence (CSC) score — a deterministic 0–100 metric that combines CIS benchmarks, CVE scanning, RBAC analysis, runtime threat detection, GitOps drift, and cluster health into a single deployment readiness number.Core Capabilities
Security Scanning
CIS benchmarks (KubeBench), CVE scanning (Grype), runtime threat detection (Falco), deprecated API detection (Pluto), and SBOM generation (Syft) — all managed by the operator.
Identity & Access
RBAC analysis across every ServiceAccount, ClusterRole, and binding. Finds overprivileged identities, unused permissions, and privilege escalation paths.
Attack Path Analysis
Maps privilege escalation graphs showing how a compromised pod could reach cluster-admin. Prioritizes the shortest, highest-impact paths.
Compliance Automation
Maps controls to SOC 2, PCI DSS, HIPAA, FedRAMP, and EO 14028. Continuous evidence collection with exportable reports.
GitOps Intelligence
Deep integration with ArgoCD and Flux CD. Tracks sync status, drift, health, and sync history without requiring API tokens.
Fleet Management
Cross-cluster intelligence correlates risks, compares configurations, and detects patterns across your entire fleet.
Signed Supply Chain
The Helm chart, the operator and agent images, and a CycloneDX SBOM for each are cosign-signed (AWS KMS). Verify exactly what you run before it reaches a cluster.
AI Risk Analysis
Plain-English explanations and prioritized fixes for findings, powered by Claude on Amazon Bedrock — with optional self-hosted NVIDIA NIM.
How It Works
Install the operator
One command deploys the ChangeGuard operator to your cluster. The operator manages the data collector, security scanners, and runtime detection — all from a single ChangeGuardAgent custom resource.
Scans run automatically
KubeBench runs CIS benchmarks every 6 hours. Grype scans container images for CVEs every 4 hours and on every new deployment. Falco monitors syscalls on every node in real time. The collector pushes cluster state snapshots every 10 seconds.
Quick Install
Quickstart
Get running in 60 seconds
Operator Installation
Install via Helm or the one-liner
CRD Reference
Full ChangeGuardAgent spec reference
API Reference
Integrate with your CI/CD pipeline