Explore the data collected, persistence, and privacy practices of the TestDriver Agent.
The TestDriver Agent is open-source and available on NPM. You can browse the source code to see all the data collected and understand how everything works.
The TestDriver Agent doesn’t contain any AI models within it. Instead, it uploads desktop context to the TestDriver API, which uses that context to make decisions about what actions to perform.
The TestDriver API leverages OpenAI models behind the scenes. For more information about OpenAI and its privacy practices, visit their privacy center.
During execution, the TestDriver Agent uploads the following information to the API:
TestDriver frequently captures screenshots of the desktop to provide the AI with decision-making context. These screenshots are uploaded to the API but aren’t persisted in the database.
Important: TestDriver can’t operate without visual context. don’t install TestDriver if you are uncomfortable with capturing images of the desktop.
Information about the open windows on the desktop is reported using the active-window
module. This includes details such as the application name and window titles.
System details, such as operating system, CPU, and memory, are reported using the systeminformation
module.
The prompts you input into TestDriver are uploaded to the API and stored in a database. This data is used to provide the AI with a history of context for better decision-making.
When running testdriver init
, you will be asked if you would like to share additional analytics. Sharing usage analytics is opt-in, and no extra data will be collected unless explicitly enabled.
To disable additional analytics, set the TD_ANALYTICS
environment variable to false
:
While the TestDriver Agent is free to use, TestDriver reserves the right to rate limit or restrict usage by IP address for any reason.
Explore the data collected, persistence, and privacy practices of the TestDriver Agent.
The TestDriver Agent is open-source and available on NPM. You can browse the source code to see all the data collected and understand how everything works.
The TestDriver Agent doesn’t contain any AI models within it. Instead, it uploads desktop context to the TestDriver API, which uses that context to make decisions about what actions to perform.
The TestDriver API leverages OpenAI models behind the scenes. For more information about OpenAI and its privacy practices, visit their privacy center.
During execution, the TestDriver Agent uploads the following information to the API:
TestDriver frequently captures screenshots of the desktop to provide the AI with decision-making context. These screenshots are uploaded to the API but aren’t persisted in the database.
Important: TestDriver can’t operate without visual context. don’t install TestDriver if you are uncomfortable with capturing images of the desktop.
Information about the open windows on the desktop is reported using the active-window
module. This includes details such as the application name and window titles.
System details, such as operating system, CPU, and memory, are reported using the systeminformation
module.
The prompts you input into TestDriver are uploaded to the API and stored in a database. This data is used to provide the AI with a history of context for better decision-making.
When running testdriver init
, you will be asked if you would like to share additional analytics. Sharing usage analytics is opt-in, and no extra data will be collected unless explicitly enabled.
To disable additional analytics, set the TD_ANALYTICS
environment variable to false
:
While the TestDriver Agent is free to use, TestDriver reserves the right to rate limit or restrict usage by IP address for any reason.