The delegator test framework provides a self-test framework (environment) for OpenHarmony applications. Using this framework, you can start an ability, schedule its lifecycle, listen for its state changes, run a shell command, and print the test result.
## Constraints
The APIs provided by the test framework can be used only in the test HAP. They take effect only after the test framework is started by running the **aa test** command or using the Integrated Development Environment (IDE).
## Starting the Test Framework
The test framework can be started by running the **aa test** command or using the IDE.
### Running aa test
You can run the **aa test** command to start the test framework. You can specify the **TestRunner** to be used and the package name or module name of the HAP where the **TestRunner** is located.
| -p | Yes | Package name of the HAP where the **TestRunner** is located. This parameter is used by the FA model. |
| -m | Yes | Module name of the HAP where the **TestRunner** is located. This parameter is used by the stage model. |
| -s unittest | Yes | Name of the **TestRunner** to be used. The TestRunner name must be the same as the file name. |
| -w | No | Timeout interval of a test case, in seconds. If this parameter is not specified, the test framework exits only after **finishTest** is invoked.|
| -s <key><value> | No | It can be any parameter in the key-value format. The entered parameters can be obtained in key-value mode through **AbilityDelegatorArgs.parameters**. For example, in **-s classname myTest**, **classname** is the key and **myTest** is the value.|
### Using the IDE
For details about how to start the IDE, see [IDE Guide](https://developer.harmonyos.com/en/docs/documentation/doc-guides/ohos-openharmony-test-framework-0000001263160453#section1034420367508).
## Introduction to TestRunner
**TestRunner** is the entry class of the test framework test process. When the test process is started, the system calls related APIs in **TestRunner**. You need to inherit this class and override the **onPrepare** and **onRun** APIs. When creating an application template, the IDE initializes the default **TestRunner** and starts the default **TestAbility** in the **onRun** API. You can modify the test code of **TestAbility** or override **onPrepare** and **onRun** in **TestRunner** to implement your own test code. For details, see [TestRunner](../reference/apis/js-apis-testRunner.md).
## Introduction to AbilityMonitor
**AbilityMonitor** is provided by the test framework for binding to and listening for abilities. You can use **AbilityMonitor** to bind to an **Ability** instance and add **AbilityMonitor** to the listening list. After an ability is bound, the creation and lifecycle changes of the ability will trigger the related callback in **AbilityMonitor**. You can test and verify the ability in these callbacks. For details, see [AbilityMonitor](../reference/apis/js-apis-application-abilityMonitor.md).
**AbilityDelegator** is a main function class of the test framework. It provides the functions of starting an ability, obtaining an **Ability** instance, scheduling the ability lifecycle, listening for the ability state, and printing test results.
**AbilityDelegator** provides APIs to display and schedule the ability lifecycle and supports the foreground and background. It works with **AbilityMonitor** to listen for the ability lifecycle. For details, see [AbilityDelegator](../reference/apis/js-apis-application-abilityDelegator.md).
### Running a Shell Command
**AbilityDelegator** provides APIs to run shell commands. You can run a shell command in the test code. This feature takes effect only in the test environment.