Ship faster. Ship from anywhere.
One pipeline language. Runs on GitHub Actions, GitLab CI, Jenkins, Forgejo, Bitbucket, CircleCI, Drone, Argo, Buildkite — your choice. Drive it from the dashboard, the CLI, or the editor insidevxcli tui.
Bring your CI provider — we drive deploys, observability, and rollback
From push to verified prod
Six stages — fan-out tests, signed images, canary, auto-rollback
Same pipeline shape no matter which provider runs the steps. Test shards parallelize automatically; SBOM and CVE scans gate promotion; canary metrics decide whether to proceed.
One dashboard · every provider
Filter by provider, repo, branch, or status — one click into logs
Stop alt-tabbing between GitHub Actions, GitLab CI, Jenkins, and Forgejo. Every run lands in the same dashboard, with the same status pills, same log viewer, and the same rerun / rollback / promote actions.
- Live status across every connected provider, refreshing in real time
- Click a row to stream the build log — no provider hop required
- Filter by provider chip, search by branch, or pin per-repo views
- Drives the same actions from the CLI: vxcli cicd build list --watch
| Provider | Pipeline · Branch | Status | Duration | Actor | Age | |
|---|---|---|---|---|---|---|
studio-web main | running | 4m 12s | joel | now | ||
platform-api-rs feat/agents | ok | 6m 38s | mira | 2m ago | ||
enterprise-payments release/2.14 | ok | 14m 02s | sam | 11m ago | ||
data-pipeline main | ok | 3m 41s | joel | 23m ago | ||
mobile-app-ios main | fail | 8m 04s | mira | 38m ago | ||
fintech-ledger main | ok | 4m 20s | sam | 1h ago |
One pipeline · four runtimes
Define once with vxcli — export to any provider
The CLI is the source of truth: vxcli cicd pipeline export --provider …emits the GitHub Actions YAML, GitLab CI YAML, or Jenkinsfile that runs the exact same stages. No copy-paste between syntaxes; no duplicated logic.
- Stages, parallelism, and strategy live in one declarative pipeline
- Provider-specific YAML stays generated — never hand-edited
- Pipeline diff in PRs shows the change, not the YAML noise
- Same vxcli runs in every CI image — predictable behaviour
1# Same pipeline, four ways. Start with the CLI — it generates the2# provider-specific YAML for you on the way out.3 4# 1) Connect a Git provider (one-time)5vxcli cicd git connect github # OAuth device flow, no copy-paste6vxcli cicd git connect gitlab7vxcli cicd git connect forgejo --url https://forge.acme.internal8 9# 2) Create a pipeline against any connected repo10vxcli cicd pipeline create \11 --name studio-web \12 --git-repo prodxcloud/studio-web \13 --branch main \14 --on push,pull_request \15 --stages build,test,scan,deploy,verify \16 --deploy kubernetes:prod-eu \17 --strategy canary:25,sticky=10m,rollback=auto18 19# 3) Trigger / watch / debug20vxcli cicd pipeline trigger studio-web --branch main21vxcli cicd build list --pipeline studio-web --status running22vxcli cicd build logs <build-id> --follow23vxcli cicd build show <build-id> --json | jq '.stages[].duration_ms'24 25# 4) Promote between environments with one command26vxcli cicd pipeline promote studio-web --from staging --to prod --canary 25Generate the pipeline from a goal
vxcli cicd pipeline new "ship studio-web with canary 25%" — emits the stages and the provider YAML.
Production-grade by default
Everything a CI/CD platform should ship with on day one
Multi-provider runtime
GitHub Actions, GitLab CI, Jenkins, Forgejo, Bitbucket, CircleCI, Drone, Argo CD, Buildkite — drive them all from one pipeline.
Parallel test sharding
Auto-shard your test suite across N runners. Re-run only failed shards. No more 40-minute test stages.
SBOM + CVE gates
Trivy / Grype scans run inline. High-severity findings block the deploy stage; SBOMs are signed with cosign.
Blue-green + canary
Canary 5%/25%/100% with sticky windows and SLO-aware promotion. Auto-rollback when error rate or p95 regress.
Vault-backed secrets
Workspace secrets resolved at runtime — no env-var gymnastics, no secrets in Actions logs.
One-click rollback
Every deploy keeps the last N images warm. Rollback is a button — or vxcli cicd pipeline rollback.
Cache + reuse
Dependency, image-layer, and test caches across runners and providers — typical 3-5× speedup on repeat builds.
Rooms + alerts
Build status, canary signals, and rollback events post into your VxCloud rooms with one click into logs.
AI failure triage
Failures fan out a vxcli chat session with the diff and stack trace pre-loaded — likely-cause hint included.
Real workloads, measured
Production traffic — across every provider
Aggregated from 30 days of customer pipelines on GitHub Actions, GitLab CI, Jenkins, Forgejo, Bitbucket, CircleCI, Drone, and Argo CD.
Where this earns its keep
From single-repo SaaS to monorepo platform teams
Single-repo SaaS shipping daily
GitHub Actions runs the build, vxcli drives the deploy — canary, verify, rollback, and a chat-room update with one config.
Mixed legacy + greenfield estate
Jenkins keeps building the JVM monolith; GitHub Actions runs the new microservices. One dashboard, one rollback button.
Sovereign / air-gapped deploys
Forgejo + on-prem Jenkins agents inside the VPC; vxcli runs the same pipelines without leaving the boundary.
Monorepo with sharded tests
Detect changed packages, parallelize tests across 16 shards on GitLab, only deploy the services that actually changed.
Provider migration in flight
Run the same pipeline on the old provider and the new one in parallel for two weeks — switch when both are green.
AI-assisted release ops
Failures auto-spawn a vxcli chat session with the diff loaded, suggested fix attached, and a one-click rerun.
Bring your CI. Keep your repo. Get the rest for free.
Connect GitHub, GitLab, Jenkins, Forgejo, or Bitbucket — and we drive deploys, observability, rollback, and AI failure triage from one place.