The Kubernetes copilot that runs inside your cluster.
Clu is an agentic AI for Kubernetes operations — troubleshooting, monitoring, scaffolding, and IaC generation. One flat hourly rate on AWS Marketplace; every capability — observability, writes, AWS context, developer onboarding — included. Installed via Helm against your EKS cluster. Runs against your own Bedrock. No phone-home, no SaaS dependency.
AWS Marketplace exclusive · Helm install · $0.10/hr per pod-hour (~$72/cluster/month)
Observability, writes, cloud context, developer onboarding.
Every capability is included at the single flat AWS Marketplace rate. Turn each one on per-cluster via Helm values; billing doesn't change based on which ones you have on.
$0.10/hr per pod-hour on AWS Marketplace (~$72/cluster/month at 24/7) — see pricing.
Always-on troubleshooter. Scans, models, and explains your cluster.
- Read-only Kubernetes observability
- Knowledge graph of resources & conventions
- Prometheus + CloudWatch integration
- Scheduled health reports + chat UI
Approval-gated writes — apply, scale, install, upgrade.
- k8s apply, scale, restart, cordon
- Helm install / upgrade / uninstall
- Dry-run + UI approval on every write
- Hash-chained audit log
Maps cluster workloads to AWS — IAM, services, cost.
- IAM + IRSA mapping & verification
- RDS, ElastiCache, S3, ECR, Secrets discovery
- Cost Explorer + networking topology
- Terraform / IaC scaffolding
Self-serve onboarding for developers — first PR ships in minutes.
- Golden-path scaffolding catalog
- Manifest, IRSA, ExternalSecret emission
- Approval-gated developer change sets
- Full-stack IaC bundles
Everything stays in your cluster. Safety is the default.
In-cluster, private by design
Clu deploys into its own namespace and talks only to the APIs you scope via IRSA. No data crosses the cluster boundary.
Bring your own model
Amazon Bedrock by default (Haiku for scheduled tasks, Sonnet for chat). Swap in any OpenAI-compatible endpoint for non-AWS clouds or air-gapped runs.
Safety-first write operations
Every write runs dry-run first, requires explicit UI approval, and is blocked in protected namespaces. Every approval is recorded in a hash-chained audit log.
Learns your conventions
The knowledge graph scans resources, infers patterns, and detects tools (Argo, Kyverno, external-secrets) automatically — no manual catalog.
Runs in your EKS cluster
Available exclusively on AWS Marketplace, installed via Helm against your EKS cluster. Single flat hourly rate; every capability — observability, writes, cloud context, developer onboarding — included. Public-registry distribution (GHCR) and additional clouds are on the roadmap.
Observability you already have
Prometheus, CloudWatch, and metrics-server are auto-detected and wired in. No new pipes to configure.
Subscribe on Marketplace, install via Helm, run.
Clu is available exclusively on AWS Marketplace, installed via Helm against your EKS cluster. Single flat rate — $0.10/hr per pod-hour — every capability included. AWS handles billing automatically: Clu calls MeterUsage once an hour while the pod is running, then stops when the pod stops.
# 1. Subscribe on AWS Marketplace — $0.10/hr per pod-hour, every capability included
# 2. Create the IRSA role — paste-ready policies in our docs
# 3. Authenticate Helm against the Marketplace ECR registry:
aws ecr get-login-password --region us-east-1 \
| helm registry login --username AWS --password-stdin \
709825985650.dkr.ecr.us-east-1.amazonaws.com
# 4. Install the chart:
helm install clu-ops-agent \
oci://709825985650.dkr.ecr.us-east-1.amazonaws.com/cloudology/clu-ops-agent \
--namespace clu-ops --create-namespace \
--set serviceAccount.annotations."eks\.amazonaws\.com/role-arn"=arn:aws:iam::ACCOUNT:role/clu-irsa
# 5. Port-forward or wire ingress
kubectl port-forward -n clu-ops svc/clu-ops-agent 8080:8080
Install Clu in your cluster today.
Available exclusively on AWS Marketplace. Start free with Core, layer on writes or cloud context when you're ready.