Guard rails come standard.

Every app Lucity builds from your code is checked for leaked keys on each deploy, and for known vulnerabilities for as long as it runs. Every workspace is walled off from the next. Nothing to install and nothing to configure, on every plan.

  • On every plan, the trial included
  • Never holds a release back
  • Open source, so you can read how it works
Leaked keysEvery deploy reads your repository and its history, and tells you which keys still work.
VulnerabilitiesApps Lucity built are rechecked every two hours while they run, not only when you ship.
IsolationApps in one workspace cannot reach the apps or databases of another.

Leaked keys,
caught on every deploy.

Each deploy reads your repository for passwords, API keys and tokens, next to the build. What it finds lands on the Security tab of the service, with the keys that still work at the top.

Rotate first.

Deleting the file does not take a key out of your history. Rotating it is what makes it harmless, and the finding stays for as long as the commit does.

The whole history, then what is newThe first deploy reads every commit. After that, only the commits since the last scan.
Exactly whereFile, line, commit and author for every finding, with the value masked.
Keys that work, firstA read-only question to the key's own provider separates live keys from old test fixtures.
Never in the wayThe release ships either way. You decide what to fix, and when.

Vulnerabilities,
checked while your app runs.

Lucity rechecks every running app it built, every two hours. A vulnerability announced tomorrow shows up on the app you deployed last month.

It keeps looking after you ship.

A check at build time goes quiet the moment the build is done. This one runs against what is live, so the list changes when the world does.

web, deployed once, four weeks ago
deployedopenssl fix announcedtoday

Rechecked every two hours since. No new deploy was needed for the finding to show up.

What sits under your code, tooYour dependencies and the base image below them, the system packages a look at your package files never sees.
The fix, not only the problemEach finding names the version that fixes it, or says that there is none yet.
Worst firstCritical and high lead the list, each marked as base image or dependency.
Where you see itOn the Security tab of each service. Alerts by email are not built yet, and prebuilt images are not scanned yet.

Walls,
between everything.

Lucity runs code from many customers on shared machines, so isolation is the first design rule and not a setting you find later.

Workspaces are walled offApps in one cannot reach the apps or databases of another.
Environments are kept apartThe private network ends at the environment, so staging cannot reach production over it.
No special privilegesRoot inside your app is not root on the machine, and apps get no access to the platform's controls.
Private databasesReachable from your own services only, unless you open one to the addresses you list.

And when something goes wrong anyway: a database restores to any second of the last 30 days, a release rolls back to the one before, and a broken deploy never replaces a working one.

Agents work inside the same rails.

Let a coding agent deploy, debug and read logs for you. The Lucity tools it gets are built so that it can do the work without holding your secrets.

Names, never valuesThe tools return which variables exist and what they point at. They never return a value or a database password.
Logs come back maskedValues Lucity knows, of eight characters or more, are replaced before the agent reads a line.
It asks, and it cannot deleteIt asks before it creates anything paid. Removing a project, service or database takes a person.

This covers what the Lucity tools return. An agent that reads a .env file on your laptop sees what is in it. How the plugin handles secrets

Answers for your security review.

The short version for whoever reviews your vendors. Where we do not have something yet, it says so.

Last reviewed 12 November 2026

AreaAnswerWhere to check
Where data lives
Apps and databases run in Germany. Buckets and database backups are stored in France. Lucity is a Swiss company.
What leaves the EU
Requests reach your apps through a global edge network, at a location near each visitor. Platform logs and metadata can be analyzed with an AI provider in the United States, under standard contractual clauses and without training on them.
Sub-processors
Four, each listed with its purpose and location. You get 30 days of notice before we add or replace one.
Isolation
Workspaces cannot reach each other. The platform runs on machines of its own. Customer apps run as unprivileged containers on shared machines, not in a virtual machine each.
In transit
HTTPS on every domain, with plain HTTP redirected. Traffic between our machines is encrypted. Public database connections are encrypted all the way to the database.
Access
Two roles per workspace, admin and member. API tokens carry a role, are shown once and can be revoked at once. Deploys from GitHub Actions need no stored token.
Not there yet
Single sign-on, an audit log of what members did, custom roles and permissions per environment. Everyone in a workspace can see and change every environment.
Backups
Every database write is archived as it happens and can be restored to any second of the last 30 days. Backups are stored with a different provider than the databases.
Your code
A scan for leaked keys on every deploy, and a vulnerability recheck every two hours, for apps Lucity builds. Prebuilt images are not scanned yet.
Breaches and deletion
We notify you without undue delay when a breach affects your data. Data of a deleted project or account is removed within 30 days.
Certifications
None of our own yet. Apps and databases run in Hetzner data centers in Germany, which are certified to ISO/IEC 27001:2022.
Reporting a vulnerability
Email security@lucity.cloud, not a public issue. We acknowledge within 2 business days and follow coordinated disclosure, with up to 90 days for a fix.

What reviewers ask us

Still wondering about something? Ask us

No. The scan for leaked keys runs next to the build and reports what it found. The release ships either way, and a failed build never replaces the version that is already running.

Only to ask the key's own provider whether it is valid, with a read-only request. Found keys are stored masked, and only members of your workspace see the findings.

A finding belongs to a commit, so it stays while that commit is in your history. The mark that a key still works is set when the key is found and is not checked again on later deploys yet. Rotating the key is what makes it harmless.

Every app Lucity builds from your repository, in every environment. Prebuilt images from public registries are not scanned yet.

Requests reach your apps through a global edge network, and some platform logs and metadata are analyzed with an AI provider in the United States under standard contractual clauses. The data processing agreement lists every sub-processor.

Yes. Lucity is open source, so they can check how scanning and isolation work instead of taking our word for it.

Hand this page to your security team.

Send us your questionnaire and we will fill it in. Or deploy something small and open its Security tab, which is the quickest way to see all of this for yourself.