finds.dev← search

// the find

devtron-labs/devtron

★ 5,508 · Go · Apache-2.0 · updated Jun 2026

The only Kubernetes dashboard you need

Devtron is a Kubernetes management platform that wraps Helm, ArgoCD, and FluxCD into a single web UI with built-in CI/CD pipelines, multi-cluster visibility, RBAC, and SSO. It targets platform engineering teams who want to give developers a self-service deployment layer without exposing raw kubectl. It is not a lightweight dashboard — it's a full platform with significant infrastructure dependencies.

- Multi-cluster, multi-tool aggregation is genuinely useful: a single UI covering Helm releases, ArgoCD apps, and FluxCD workloads across clusters reduces tool-switching for ops teams managing heterogeneous environments.

- RBAC via Casbin gives you policy-as-data rather than hardcoded role logic — fine-grained access control down to individual K8s resources is hard to build well, and Devtron doesn't make you roll it yourself.

- The Helm rollback and config drift detection features solve real day-to-day pain: comparing what you deployed vs. what's running is something vanilla ArgoCD doesn't make easy.

- Active development with a detailed CHANGELOG going back years and daily commits as of this writing — this is not an abandoned project.

- The `_ent.go` files scattered throughout the codebase (BulkUpdateRestHandler_ent.go, HelmAppService_ent.go, DeploymentPipelineRestHandler_ent.go, etc.) signal an enterprise edition split — you won't know what's missing from the open-source version until you hit the wall.

- Heavy install footprint: PostgreSQL, NATS, Dex, Casbin, and a suite of microservices just to get a dashboard running. This is not something you drop into a cluster in 10 minutes; plan for a real ops burden.

- The Wire-based dependency injection with a top-level Wire.go and WiringNilCheck.go on a codebase this large creates a contribution barrier — tracing what wires to what is non-trivial and the code gen adds a build step most Go projects don't have.

- The 'AI-Native' headline in the README is pure marketing positioning with no substance shown — the README doesn't describe what the AI teammate actually does, which is a red flag about whether it exists in any meaningful form.

View on GitHub → Homepage ↗

// want more like this?

We dig through GitHub every week and send a few repos picked for what you actually care about — each with an honest take like this one.

Get finds in your inbox → Search again →