Add one step to the CI you already run and get portable, isolated, policy-checked builds with signed provenance — same run, same result, on any CI and any cloud. Or run it standalone.
The same pipeline.yaml runs as a layer on the CI you already have, on any other CI or cloud, or on its own. Pick a path — the definition never changes.
# add one step — nothing else changes steps: - uses: actions/checkout@v4 - uses: lab-hit/labhit-action@v1 with: pipeline: .labhit.yaml
# one binary — embedded storage, no services $ labhit run → scheduling 6 stages… → sandbox ready · policy gate ok → pipeline passed in 1m38s
# the portable definition — runs anywhere engine: "1" pipeline: name: build-and-deploy # same file on any CI, any cloud, your laptop
Each stage declares what to run — an extension or a shell command. Wire dependencies with after. The scheduler builds a DAG and runs independent stages in parallel.
engine: "1" pipeline: name: build-and-deploy stages: fetch: use: source/git with: depth: 1 test: after: [fetch] run: cargo test --workspace sandbox: image: rust:1.93-slim build: after: [test] use: build/container with: dockerfile: Dockerfile scan: after: [test] use: scan/trivy deploy: after: [build, scan] use: deploy/kubernetes gate: approval: required
The engine ships with zero built-in integrations. Every capability is an extension you install by name — source/git, build/container, deploy/ethereum. All sandboxed, all composable.
Browse the extension interfaceBrowse, install, and publish extensions across 15 categories. Build in any language that compiles to WASM.
Solidity builds, Foundry tests, Ethereum & Solana deploys, gas audits, contract verification.
PyTorch training, ONNX builds, model validation gates, GPU containers, inference deployment.
SAST, DAST, dependency scanning, secret detection, license gates, benchmark checks.
Any language that compiles to WASM. Publish free, or sell your own.
Same config. Same extensions. Same results. Scale up only when you need to.
Run on your laptop. Zero infrastructure. Start with labhit run and watch your pipeline execute.
Deploy to a server. Same binary, same config. Add a database for persistence and your team is running.
Scale horizontally. High availability, distributed scheduling, same extensions. The config never changes.
The config never changes. Only the infrastructure beneath it does.
Built in the open. Shipped when ready. Follow our engineering milestones.
Add it to the CI you already use, or run it standalone as a single binary. Try the beta, or read the open spec.