Security

Vibecode Security

We build the platform so that everything is locked down by default. The server your AI agent creates is invisible from the internet. By default, only you have access to your app. Keys are issued with limited permissions. You stay the owner of your data and decide who can see what.

The essentials

1

Two-factor sign-in

One step blocks 99% of account-takeover attempts. Takes a minute to turn on.

2

The unified key is like a safe key

It opens everything you have access to. Never paste it into code, never share it in chats.

3

Public mode, used deliberately

It means full access from the internet with no sign-in. Do not enable it unless you understand the consequences.

Who is responsible for what

Three parties are involved when you work with the platform. Each has its own zone. This helps clarify what we take on and what depends on you.

Platform

Vibecode

  • Infrastructure security
  • Data protection in transit and at rest
  • Isolation of apps from one another
  • Issuing and revoking keys
  • Access and action logs
AI tool

Claude Code, Cursor and others

  • Generating the app code
  • No guarantee that the code is secure

This is a third-party service. Code from any AI is worth reviewing - that is common industry practice, not a Vibecode quirk.

You

User

  • Which data to entrust to the app
  • Which permissions to grant the AI agent and keys
  • Who gets access to the app
  • Reviewing the result before launch
  • Password, two-factor sign-in, account protection
  • Agreements and the law regarding your customers’ data

What we take on

Works without your involvement and protects you by default.

Black Hole servers

By default every server starts in non-public mode - all ports closed, no access from outside. Only those you grant access to can reach the server, and only through an encrypted tunnel.

Access only for you

Every app is available only to its owner by default. Widening access is step-by-step: employees, department, the whole Bitrix24 account. Every change is logged.

Data protection

Data is protected in transit over a secure channel and at rest. Keys to third-party services are stored encrypted.

Trusted infrastructure

Vibecode servers run on Bitrix24 Cloud. Vibecode is part of the Bitrix24 ecosystem.

Secure sign-in

Sign-in is designed so that secret session data never reaches the visitor’s browser. Attempts to impersonate someone are cut off right at the entry point.

Flexible access keys

You can limit permissions, set an expiration date, turn on read-only mode (where nothing can be changed or deleted), and revoke a key at any time.

The essentials

The unified Vibecode key is like a safe key

To let an AI agent start building an app for you, you issue it the unified Vibecode key and pass it to an external tool. In effect this key is the master pass to everything you have access to. Whoever holds the key controls the access. Treat it more carefully than an ordinary password.

How to handle it

  • Save it when you create it - the key is shown only once.
  • Store it in a safe place known only to you.
  • Never paste the key into code or public resources.
  • Do not send it over messengers or email.
  • One key, one task. That makes it easier to revoke.
  • Rotate and revoke: a new key from time to time, and disable the old ones.

What a stolen key threatens

A stolen key means access to your data. Through it an attacker can run AI requests and burn through your balance - sometimes before you even notice.

If a key might have ended up in the wrong hands

Revoke it in the interface, create a new one, update the apps that use it, and review the request log for unfamiliar addresses.

What depends on you

Six simple actions. No programming skills required.

1

Protect your account

Most accounts are lost not because of holes in platforms, but because of weak passwords and stolen credentials. We start here.

  • Turn on two-factor sign-in (2FA). Signing in will require a one-time code. Even if your password is stolen, no one gets in without the second step. This applies to both of your accounts - in Bitrix24 and in Vibecode. An administrator can make 2FA mandatory for the whole company.
  • Set a strong, unique password. At least 10–12 characters, no names or birth dates, and not the same as on other sites.
  • Sign out of your account on other people’s devices. The "Sign out" button is in the menu in the top-right corner.
2

Grant AI agents the minimum permissions

This is the most important rule of vibe coding, and here is why.

  • An agent gets exactly the same permissions as your account. Imagine you handed someone a key ring: they can open every door those keys fit.
  • Minimum permissions. If you need to read deals, do not grant the right to delete them.
  • Read-only mode. If an app only needs to display data, turn on READONLY - then it physically cannot change anything.
3

Do not feed the AI more than necessary

AI tools are third-party services. Before pasting anything into a prompt, think about what exactly you are sending.

  • Do not paste secrets - passwords, keys, access tokens.
  • Be careful with customers’ personal data. If the task does not need real data, use test data.
  • Remember your own key (BYOK). If you connect your own key to OpenAI, Anthropic, Google and others, you pay the provider directly. A third-party AI service may use the content you send it under its own rules - read the specific provider’s terms in advance.
4

Check the app for vulnerabilities before launch

Good news for those who cannot read code: you can delegate the review to an AI. The bad news - you cannot delegate it to the same agent that wrote the app: it tends to "justify" its own work and may convincingly say everything is fine even when it is not.

  • Run a separate, clean review. Open a new session (or a different AI tool / a different model) and grant the agent read-only access. Let it act as an independent security auditor.
  • Use a clear prompt: "Check this app and its server for common leaks and holes. Find and list in plain language, without jargon: 1) open endpoints that work without authorization; 2) passwords, keys or tokens hard-coded right into the code; 3) which data the app accesses and whether it can change or delete more than it should; 4) who can open the app right now. For each item, say whether it is dangerous and what to fix."
  • Check the settings yourself - this needs no code. Look in the interface: the key has no excess permissions, the access policy is definitely not "Public" by mistake, the app behaved as expected on deliberately test data.
  • For sensitive tasks, bring in a human. If the app handles payments, medical or customers’ personal data, do not rely on an AI review alone - ask a specialist or a certified Bitrix24 partner to look at the result. That is not a sign of weakness but a normal, mature practice.

If you are not sure, it is always best to entrust the app’s security audit to an information-security specialist or a company that provides such services.

5

Open access deliberately

When the app is ready, decide who should use it and set access on the "no more than needed" principle.

6

If you publish the app for others

Planning to put the app in a catalog or let others use it? Then you become its author and must attach your own terms of use and privacy policy, compliant with applicable law, so that a person cannot start using the app without agreeing to the terms.

Public mode

What happens if you turn on "Public"

The most open access mode is called "Public". Before enabling it, it is important to understand exactly what happens.

When you switch to "Public"

  • The app gets an ordinary internet address available to anyone.
  • Sign-in is no longer required - no password, no invitation, no account.
  • Everything the app can show and do becomes available to everyone.
  • Bots constantly scan the internet and find open apps within an hour. Your app may, for example, end up in search-engine results.

Therefore

The server never switches to public mode on its own - only after your explicit confirmation, and every such change is recorded in the log.

Turn on "Public" only when you truly need to - for example, for a publicly available form or landing page that has no sensitive data and no write access to your Bitrix24 account. If in doubt, do not enable it. Opening access later is always easier than dealing with the aftermath of a leak.

What is important to know about AI

These things apply to any AI tool, including Vibecode. There is no way around them.

AI acts on your behalf

Whatever it does counts as your actions. If, because of overly broad permissions or an unclear prompt, the agent does something undesirable, you are the one responsible. So keep permissions minimal and stay in control.

AI can be tricked

An attacker can slip a hidden command inside a page, document or email that the AI reads. This is called prompt injection: the model sees an "outside" instruction and sometimes follows it. The defense is simple - give the agent minimal permissions and do not connect unnecessary external sources.

AI makes mistakes

Generated text and code may not match the facts - the AI can "invent" a feature or get the logic wrong. Review the result and do not trust important decisions with legal or financial consequences to AI alone - the final word stays with a human.

BitrixGPT

About our model, separately

The platform’s built-in BitrixGPT model runs within our perimeter and does not use your data for its training. Your prompts are processed only to answer the specific request and do not enter the training set.

If you connect third-party AI services through your own key (BYOK), that service’s policy is on the provider’s side. The terms of use are worth reading with the specific service.

So you notice what matters in time

Some risks are removed not by a setting but by attention - we try to help you keep it.

Confirmation in the interface

Risky actions require confirmation. Turning on "Public" mode, issuing a new key with broad permissions, changing two-factor sign-in settings - in each case we show exactly what will happen and ask you to confirm.

Emails about important events

We try to notify you by email about critical operations and suspicious sign-in attempts. If you receive an email about an action you did not take, that is a reason to check your account right away and change your password if needed.

Checklists - five minutes before launch

Copy them, go through the items, and you will head off most problems in advance.

Before launching the app

  • I ran an independent AI review for errors (in a separate session, read-only access).
  • I checked in the interface: the key has no excess permissions.
  • The unified Vibecode key is stored in a safe place, not in code and not in chats.
  • The access policy matches the task. Public mode is not enabled by mistake.
  • The app was tested on deliberately test data.
  • For sensitive data, a specialist or partner reviewed the result.

Account and team security

  • Two-factor sign-in is on (mandatory for the whole company where possible).
  • The password is strong and unique.
  • Employee permissions are granted on the "minimum necessary" principle.
  • It is configured who can invite new users.
  • From time to time I review and revoke unused keys.
  • The team knows what phishing looks like (see below).

How to recognize phishing

Phishing is an attempt to extract a password or data through fake emails and sites. It is the most common and effective way to break in - and it is easy to spot once you know the signs.

!

A similar but not quite right address

For example, bltrix instead of bitrix, a zero instead of the letter "O", an extra hyphen.

!

Urgency and pressure

"Urgent", "or access gets blocked today", "5 minutes left".

!

A request to follow a link and enter your password

Real services do not do this. Open the site manually via a bookmark.

!

Suspicious attachments

Files like .exe, .scr or with a double extension such as document.pdf.exe.

In doubt - do not follow a link from the email. Open the service manually via a bookmark or a search box. Confirm a request for money or data through another channel - a phone call or a work chat.

A short glossary

No jargon. If something on the page is unclear, the explanation is probably here.

2FA (two-factor sign-in)

Sign-in in two steps: a password plus a one-time code.

The unified Vibecode key

The master pass you issue to AI agents. It opens access to everything you have access to.

Key permissions (scope)

What exactly the pass is allowed to do - for example, only read the CRM.

Prompt injection

A command hidden in text that can trick the AI.

BYOK

"Bring your own key" - using your personal key to a third-party AI service.

Personal-data controller

The one who decides what data about people to collect and why (when it comes to your customers, that is you).

Black Hole servers

Vibecode servers with no public internet address - closed by default, reachable only through an encrypted tunnel.

Public mode

Access to the app from the internet with no sign-in, for any visitor.

If something goes wrong

A calm, fast plan for three cases.

You suspect your account was hacked or a key leaked

Change your password, turn on 2FA, revoke suspicious keys and sign out of all sessions. Then report it to support through the feedback form. The faster, the better.

You noticed a problem in your own app

Excess access, strange behavior, an unexpected bill for AI requests - narrow the access policy or temporarily switch the app to read-only mode, revoke excess permissions from the agent and keys, and check the access log.

You found a vulnerability in the platform itself

Please report it to us promptly through the feedback form in your account. Do not disclose the vulnerability publicly until it is fixed - for our part, we will try to respond promptly.

For help, write to us through the feedback form in your account - it is the fastest channel.