---
title: Creating a monitor
description: >-
  Learn how to create, edit, duplicate, and bulk-generate monitors from tests to
  run scheduled checks with alerts in production.
sidebarTitle: Create a monitor
lastUpdated: "2026-09-09"
---

> **For AI agents:** the complete documentation index is at [llms.txt](/llms.txt). Append `.md` to any page URL for its markdown version.

Create a monitor from the Tests workspace or directly from the Monitors list. A monitor packages a test with a schedule, device assignment, and [alert rules](/alerting/alerts) so it runs continuously in production.

## Create a new monitor

<Steps>
<Step>
Open **Monitors** and click the **Create** button, or open a test, select it in the context menu, and choose **Create Monitors**. You can also promote a run you've just reviewed: on an execution result page, click **Create a monitor** (top right) and the form opens pre-filled with the test's details.
</Step>

<Step>
Select the test you want to monitor. Once selected, the form loads with the test's variables and device requirements.
</Step>

<Step>
**About**
- Enter a name for the monitor (required; max 200 characters). By default, it uses the test name.
</Step>

<Step>
**Variables** (if the test has them)
- Fill in each variable: the test's own [variables](/tests/variables), global variables, variable groups, and owner-type device parameters.
- A value can reference a global variable or a built-in random, so every scheduled run can use fresh data.
- Owner-type variables require you to select a device for each one.
- All variables must have a value before you can save.
</Step>

<Step>
**Device**
- Select the device(s) the test will run on: SIM, Smartphone, or Web agent.
- Only appears if the test requires devices.
</Step>

<Step>
**Scheduling**
- Choose simple (every X seconds/minutes/hours) or advanced scheduling (specific days, times, timezones). See [Scheduling](/monitoring/scheduling) for every option.
- Default is every 5 minutes.
</Step>

<Step>
**Options** (optional)
- Enable auto-retry (1–3 retries) if the test fails. A warning notes that retries can be delayed when the agent's parallel capacity is full.
- Enable **Ignore errors** to mark the test as passed even if assertions fail.
- Set SMS timeout (in seconds) if the test waits for SMS.
</Step>

<Step>
Click **Create** to save the monitor. It will start running on its schedule immediately.
</Step>
</Steps>

## Edit a monitor

Open a monitor and click **Configure**. You can change the monitor's name, scheduling, devices, and options. If the monitor is paused, a dialog will ask whether to unpause it when you save.

<Info>
You cannot change the test after the monitor is created. To use a different test, duplicate the monitor and create a new one from a different test instead.
</Info>

## Duplicate a monitor

Duplication copies all settings from an existing monitor. Open the monitor's menu and click **Duplicate**, or open the monitor's detail page and click the three-dot menu.

You can rename the monitor and reassign devices before saving. The new monitor is created as a copy with all the same scheduling and variables.

## Create monitors in bulk from tests

Select one or more tests (they must all be the same type), open the context menu, and click **Create Monitors**. A modal appears:

1. Select a single device that all monitors will use.
2. Configure the schedule.
3. Click **Create X Monitors**, where X is the number of selected tests.

The system creates one monitor per test with the shared device and schedule. Each monitor uses the test's default variables and configuration.

## Form validation

Before you can save, the form checks:

- Monitor name is not empty.
- All required variables have values.
- All owner-type variables have a device assigned.
- For simple scheduling: frequency is set.
- For advanced scheduling: at least one weekday is selected, and the start date is not after the end date.
- For web tests: a browser preset is selected.

If validation fails, the form scrolls to the first error and shows a message.

<Tip>
Creating many monitors, or managing them from scripts? The CLI covers the same operations: see [Monitors and incidents from the CLI](/cli/guides/monitors-incidents).
</Tip>