Zero-config builds
Language autodetect (Node, Python, Go, Rust, Java, Ruby, PHP, .NET) renders a Dockerfile if your repo doesn't ship one. EXPOSE and per-language port conventions become your defaults.
liftwork is a plug-and-play build & deploy platform for Kubernetes — a self-hosted internal Heroku. Drop the chart into your cluster, point at a repo, and a git push becomes a deploy. On infrastructure you own.
Language autodetect (Node, Python, Go, Rust, Java, Ruby, PHP, .NET) renders a Dockerfile if your repo doesn't ship one. EXPOSE and per-language port conventions become your defaults.
Commit a liftwork.yaml at the repo root and the file wins — port, replicas, env (incl. Secret & ConfigMap refs), health probes, ingress, resources. GitOps-native.
No privileged containers, no external build farm. Cache, multi-arch, secret mounts — all the BuildKit features you'd expect, running as a Kubernetes Job inside your cluster.
Rollout watcher classifies failures into actionable diagnoses — "container is up but isn't listening on port 9999" — and bails in 9–15s instead of timing out at 600s.
~18 Prometheus metrics out of the box, OTel traces across every job, two ready-to-import Grafana dashboards (Operations + SRE) committed alongside the chart.
API, worker, dashboard, RBAC, ServiceMonitor, optional bundled Postgres + Redis. Namespace-aware values mean a single chart can isolate per-tenant installs.
Bundled Postgres + Redis subcharts mean zero external dependencies for dev. Production paths point at your managed services via externalDatabase.url / externalRedis.url.
Full setup, per-tenant install, and prod recipes are documented in docs/install.md.
# clone & bootstrap $ git clone https://github.com/P0intMaN/liftwork.git $ cd liftwork && helm dep update charts/liftwork # install with bundled postgres + redis $ helm upgrade --install lw charts/liftwork \ --namespace liftwork --create-namespace \ -f charts/liftwork/values.dev.yaml → migrate Job (alembic upgrade head) → api · worker · dashboard rolling out → ready in ~30s