Deploy from GitHub on every push
Connect a repository through the Livops GitHub App, pick a branch, and every git push to that branch builds and ships a new version, with the build log in the cabinet. A failed build leaves the running version alone, and a previous version is one click away in the Deploys list. To ship only after your tests pass, turn on Deploy after CI passes or call a deploy hook from GitHub Actions.
By Livops team, the people who build and run LivopsUpdated
Pushing to GitHub and seeing the new version live a minute later is the whole idea of a deploy platform. This page shows how to connect a repository to Livops, what happens on each push, how to roll back, how to make deploys wait for your tests, and what the same thing takes when you build it yourself with GitHub Actions and SSH to a VPS.
What you need
- A Livops account.
- A GitHub repository with your app, bot or site. GitLab works too, through a webhook in the project settings.
- Permission in GitHub to install an app on that repository, or on the organisation that owns it.
Connect the repository
- In the cabinet, click New project, then Connect GitHub. Livops is installed as a GitHub App and sees only the repositories you select.
- Choose the repository and the branch. If you pick no branch, the repository’s default branch is used.
- For a monorepo, give the folder the project lives in, for example
apps/web. - Livops detects the framework and suggests the install, build and start commands. Change them if they are wrong, then deploy.
The project gets a <name>.livops.app address with HTTPS. From then on, every git push to that branch starts a new deploy; pushes to other branches do not.
What happens on each push
- The code is fetched at the commit you pushed, and the log says which.
- Dependencies install and the build runs. The build log is on the deploy’s page in the cabinet as it runs. A build has ten minutes; one that runs past that is stopped as
timed_out, and the previous version keeps running. - The new version starts. A web app has a minute to answer on the port in
PORT. Until it does, the old version keeps serving. - If anything fails, nothing is switched. The log ends with the error and often a Tip: line with the fix, the error has a code (
build_failed,start_failedand so on) that links to its page in the docs, and the assistant button next to the failed deploy explains what to change.
A bot is the one exception to step 3: its previous version is stopped before the new one starts, because two copies polling Telegram would conflict.
To pin the commands in the repository instead of the cabinet, add livops.json to the root; it is read on every deploy, from the commit being built:
{
"kind": "web",
"build": "npm run build",
"start": "node dist/server.js",
"healthPath": "/healthz"
}
Roll back or deploy again
- Roll back. Recent versions are kept; open the Deploys list and roll back to a previous one in one click.
- Deploy again without a push. From the terminal,
livops deploy <service>deploys and follows the build until it is live or failed. A deploy hook does the same with onePOST. - From an AI client. Over MCP,
rollbackputs an earlier deploy back in service anddeployredeploys from the current source.
Deploy only after your tests pass
By default, a push is built at once. There are two ways to make it wait.
Deploy after CI passes
On a service connected from GitHub, turn on Deploy after CI passes. A push then waits in the Waiting for CI state; when that exact commit’s GitHub checks pass, the deploy starts. If a check fails, is cancelled or times out, nothing is built and the old version keeps running. A newer push replaces the one that is waiting.
A deploy hook at the end of your workflow
A deploy hook is a secret URL: a POST to it deploys the branch the service tracks. Create one in the service settings under Deploy hook — the URL is shown only once — and store it as a repository secret, under Settings → Secrets and variables → Actions, which a workflow reads through the secrets context[2]. Then make this the last step of a workflow that runs on push to main — a branches filter limits it to that branch[1] — after your tests:
curl -fsS -X POST "$LIVOPS_DEPLOY_HOOK" # the secret, passed to the step as an env variable
-f makes curl fail on an error status, so a wrong or revoked hook fails the job instead of passing quietly. The complete workflow file is in Deploy hooks and CI. If you use a hook this way, keep the push from also deploying on its own, or each push builds twice: Deploy after CI passes is the simpler choice when your checks already run on GitHub.
Get told when a deploy goes out
Under the team’s Settings → Webhooks, add an HTTPS address and Livops sends a signed JSON POST when a deploy goes live, fails, or an app stops. A Slack or Discord webhook address gets that chat’s own message instead, such as “✅ shop — new version is out” or “❌ shop — did not go out” with the reason.
By hand: GitHub Actions and SSH to a VPS
You can build the same pipeline yourself: a workflow that runs on push to main[1], a private SSH key stored as a repository secret[2], and a step that writes the key to a file and runs one command on the server:
ssh -i key [email protected] \
'cd /srv/app && git pull && npm ci && npm run build && sudo systemctl restart app'
It works, and it costs little: standard GitHub-hosted runners are free in public repositories, and private repositories get a monthly allowance — $12 of usage on GitHub Free, $18 on Pro[3]. Secrets other than GITHUB_TOKEN are not passed to workflows triggered from a fork[2], so a pull request from outside cannot read the key.
What the script above leaves to you:
- No previous version to go back to.
git pulland a restart replace the running code in place; if the new build breaks, the site is down until you fix it or check out the old commit by hand. - Downtime on every restart, unless you add a second process and a proxy that switches between them.
- No build log outside the Actions run, and no reason in plain words when it fails.
- The server itself: the operating system and its updates, the firewall, the process manager, HTTPS, and keeping the SSH key’s access narrow.
If you want root and full control of the machine, that is a fair price. If you want a push to become a running version and nothing else, a platform does those steps for you.
Deploying from GitHub on Livops
- New project → Connect GitHub, select the repository in GitHub.
- Pick the branch and, for a monorepo, the folder.
- Check the detected commands and press deploy. Watch the build log.
- Push. Every push to that branch is a new version; a failed one leaves the running version alone.
- Optional: turn on Deploy after CI passes, add a webhook for Slack or Discord, and add a
stagingenvironment to test before production — see Environments.
A static site is free; a bot is 19 credits a month and an app 39 credits. Builds count against your plan’s monthly allowance; past it, each build is 1 credit. Credits are bought in so’m through Payme or Click, with no card. See Ways to deploy, Deploy hooks and CI and Pricing.
Questions
Which branch does Livops deploy?
The one you pick when you connect the repository, or the repository's default branch if you pick none. Pushes to other branches do not deploy. You can change the branch later in the project's Settings.
Can Livops see all my repositories?
No. Livops is installed as a GitHub App and sees only the repositories you select in GitHub. For an organisation's repository, grant the Livops app access in the organisation's settings.
What happens when a build fails?
Nothing changes for your users: the version that was running keeps running. The log shows why the build failed, often with a Tip line that says what to do, and the assistant button next to the failed deploy explains it.
How do I roll back?
Open the project's Deploys list and roll back to a previous version in one click. Recent versions are kept for this. The assistant, the API and MCP clients can roll back too.
Can I deploy only after my tests pass?
Yes, two ways. Turn on Deploy after CI passes, and a push waits until that commit's GitHub checks pass. Or create a deploy hook and call it with curl as the last step of your GitHub Actions workflow.
Does a monorepo work?
Yes. When you connect the repository, give the folder the project lives in, for example apps/web. Each project in the repository gets its own folder setting.
Is GitLab supported?
Yes, through a GitLab webhook that you connect in the project settings.
Sources
- GitHub Docs — Events that trigger workflows (docs.github.com). Checked .
- GitHub Docs — Using secrets in GitHub Actions (docs.github.com). Checked .
- GitHub Docs — GitHub Actions billing (docs.github.com). Checked .
Put your project online today
Connect GitHub or drop a folder. No card is asked for.
Connect GitHubRelated
- A staging environment next to productionOn Livops, Add environment on the project page makes a staging copy of your production services, each with its own address, its own variables and its own empty database, so a test never touches production data. Promote to production then ships the exact release you tested, without rebuilding. A staging environment costs half; pull request previews are billed like regular services.
- What Livops's AI assistant does when a build failsWhen a build fails on Livops, the previous version keeps running and one button asks the assistant why. It reads the failed build's log and your project's commands, says what went wrong and how to fix it, and may offer one fix as a button that changes nothing until you press it. It never deletes, never sees variable values, and does not edit your code.
- npm ERR! code 1: why the build failed with exit code 1npm ERR! code 1 (npm error code 1 in newer npm) only says that a script npm ran exited with code 1. The real error is printed above it: usually a compile or type error, a lockfile out of sync with package.json, or a dependency whose install script failed. Find the first error line above code 1, fix that, and the build passes.
- App not listening on $PORT: why the health check failed after deployThe platform started your app, tried the port it gave it, and nothing answered in time. Almost always the app listens on its own fixed port instead of the PORT variable, listens on localhost (127.0.0.1) instead of 0.0.0.0, or crashed or was still starting when the check gave up. Listen on 0.0.0.0 and the port in PORT, and it passes.
- Deploy from Claude Code: connect Livops over MCPRun one command, claude mcp add --transport http livops https://livops.uz/api/mcp, then sign in through /mcp. From then on Claude Code can list your projects, deploy, read the build log, explain a failure and roll back, as you and only in the team you picked.
- Managed PostgreSQL hosting in 2026: the options comparedA production PostgreSQL needs backups you can restore, a connection limit your app stays under, the app in the same region, and a supported version. On Livops a PostgreSQL 15, 16 or 17 database runs inside your project, its DATABASE_URL is written into your services, and it is backed up every night; 500 MB is free, one per team, and 1 GB is 9 credits a month. Here is how Neon, Supabase, Render, Railway and a VPS compare.