Release v0.6.0
We shipped v0.6.0, and we celebrated the preview with a launch party in Gothenburg, hosted by fabled.se. Thanks to everyone who came.
It’s our biggest release so far, and most of it is about Kubernetes. Operations is now a single view of every workload across all your clusters, with restart counts, resource limits, and deployment history in one drawer.
A few highlights:
- Cluster-wide risk signals. Operations flags workloads with risky RBAC (cluster-admin, wildcard verbs, or secret access) and services exposed through a LoadBalancer or NodePort.
- Helm and GitOps aware. The agent detects Helm releases and shows ArgoCD and Flux sync status per cluster.
- One agent per cluster. Label a namespace with its team and every workload there lands in the catalog owned by the right team, with no per-workload config.
- Security in one place. An overview, a findings list, and a detail view, built from your real CodeQL, Dependabot, and secret-scanning data.
- Notifications to Slack and email, next to the in-app inbox.
- A visual refresh across Operations, Catalog, Dashboard, Insights, Forge, and Admin.
Plenty of bug fixes along the way too.
We’re taking a short vacation break and will be back mid-July to start on v0.6.1: more Git providers, Microsoft Entra ID auth, and an enhanced MCP server.
Changelog
The full changelog lives here.
In other news
- shoehorn-k8s-agents is now public and open source.
- Vote on a feature