DevOps Engineer passionate about Cloud Infrastructure, Kubernetes, and Automation. Building open-source tools for the DevOps community.
Follow for new Kubernetes operators, GitHub Actions, and OSS contribution updates.
| Project | Description | Stars |
|---|---|---|
| compress-decompress | GitHub Action for file compression/decompression |  |
| network-policy-generator | Kubernetes operator that generates NetworkPolicy from a CRD |  |
| cicd-monitoring | Production-ready CI/CD & monitoring stack for AWS/GCP/on-prem Kubernetes |  |
| kube-ctx | Single-binary kubectx/kubens replacement with per-terminal context isolation, production guards, and a built-in fuzzy picker |  |
| kube-diff | CLI that diffs Kubernetes manifests (plain YAML, Helm, Kustomize) against the live cluster |  |
| whatdies | Claude Code hook that shows what a Terraform plan or kubectl command would destroy before your AI agent runs it |  |
| helios-lb | Kubernetes controller providing bare-metal LoadBalancer IP allocation (MetalLB-style) |  |
| k8s-namespace-sync | Kubernetes controller that syncs Secrets/ConfigMaps across namespaces |  |
| git-bridge | Multi-provider Git mirror: CodeCommit, GitLab, GitHub any-to-any with webhooks & SQS |  |
Contributions to external open-source projects.
| Project | PR | Contribution |
|---|---|---|
| apache/airflow | #67675 | Add Gateway API HTTPRoute support for the API server to the Helm chart |
| elastic/elasticsearch | #151614 | S3 snapshot repository: EKS Pod Identity credential support |
| go-gitea/gitea | #38863 | Stream bundle downloads instead of writing a repo-sized temp file to disk first |
| distribution/distribution | #4995 | Return 400 NAME_INVALID instead of 500 UNKNOWN for repository names longer than 255 characters, and stop the registry dispatcher's early returns from writing the error response twice |
| pulumi/pulumi | #24995 | Redact secret values as [secret] in invalid input diagnostics unless --show-secrets is passed, so a provider Check failure, the pulumi import warning or a Construct or Call input error no longer prints them in plaintext |
| prometheus-operator/prometheus-operator | #8728 | Add spec.retentionPercentage to the Prometheus CRD for percentage-based TSDB retention (Prometheus >= v3.11.0) |
| projectcalico/calico | #13979 | Accept the full 4-byte AS number range (RFC 4893) in the BGPConfiguration, BGPPeer and BGPFilter CRD schemas |
| open-telemetry/opentelemetry-collector-contrib | #49146 | Add opt-in TLS support to the memcached receiver |
| nginx/nginx-gateway-fabric | #5392 | Add GEP-713 Programmed status condition to custom policies |
| getmoto/moto | #10062 | Add EC2 snapshot tiering APIs (archive / describe-status / restore) |
| anchore/grype | #3519 | Add vulnerable version ranges to the CycloneDX output format |
| tektoncd/pipeline | #10548 | Filter ResolutionRequests by the resolver's own label selector on leader promotion, so one resolver no longer fails requests owned by another |
| fluxcd/flux2 | #5975 | Add an opt-in version-file input to the Flux CLI GitHub Action |
| k3s-io/k3s | #14904 | Close idle client connections on the supervisor: set IdleTimeout to 90 seconds, the value kube-apiserver uses, so a client that went away without a FIN no longer leaves its connection and goroutines on the server until restart |
| vectordotdev/vector | #25607 | Add a host_metrics temperature collector via sysinfo Components |
| hashicorp/terraform-provider-aws | #50463 | Fix aws_prometheus_scraper update failure on scrapers with a CloudWatch destination by omitting unchanged destination from update requests |
