8. The MultiConnect CLI
You need the developer's approval before each of these commands (Section 2).
| Command | Changes anything? | What it does |
|---|---|---|
multiconnect status |
No | State, build version, rollout, containers and health |
multiconnect info |
No | Connection details. Summarise them |
multiconnect analytics |
No | Cost per hour, usage against the plan, per-container resources |
multiconnect balance |
No | Credits and spend |
multiconnect deploy |
Yes, production | Builds, zips, uploads and rolls out with current settings |
multiconnect deploy --config |
Yes | Needs a real terminal, so the developer runs it |
multiconnect init |
Writes multiconnect.json |
The developer runs it |
deploy options: --tag <label>, --skip-build, --no-rollback, --no-wait, --timeout <minutes>, --yes. The CLI asks nothing in a non-interactive shell, so the developer's yes in chat is the only check.
| Exit code | Meaning | Tell the developer |
|---|---|---|
0 |
Live | It is deployed. Old containers may still be draining, which is normal |
1 |
Failed or refused | Quote the error. If it says rolled back, the previous build is still serving |
2 |
Still rolling out after --timeout |
Not a failure yet. Offer to run status |
multiconnect.json links a folder to one deployment. You may read DeploymentName, DeploymentType, IsMasterWorker, BuildPath, BuildCommand, WorkerBuildPath, WorkerBuildCommand and RollbackOnFailure. To describe a request, read those fields. To set up a pipeline, use MC_ACCESS_KEY and MC_ACCESS_SECRET from the CI system's secret store, and say plainly in your summary that the pipeline deploys.
If the folder has no multiconnect.json, the developer sets it up: create a key with Allow deploys (CLI) on under Access Keys, attach it on the deployment's page, then run multiconnect init.

