Browser agent
Also called computer-use agent, web agent
A browser agent is an AI model that operates a web browser: it reads the page, decides on an action such as a click or a keystroke, performs it, and repeats until the task is done.
The loop is simple to describe. The agent takes a screenshot or reads the page structure, the model chooses the next action, the browser carries it out, and the new state goes back to the model.
What makes one agent more useful than another is everything around that loop: a session that keeps cookies, an inbox that can receive a verification email, the ability to pause for a human when it meets a CAPTCHA, and a record of what it did.
Where Stunt Double fits
Actors are browser agents with a persona, knowledge and an inbox of their own. Runs happen in hosted browsers, or on your own machine or network through the Mac app or a self-hosted worker.
Related terms
- Agent experience (AX)Agent experience is how well an AI agent can use a product on someone’s behalf: find what it needs, understand the page, and finish the task without a person stepping in.
- AI user testingAI user testing sends AI agents with defined personas through a product in place of recruited participants, and reports what each one did and where it got stuck.
- Model Context Protocol (MCP)The Model Context Protocol is an open standard for connecting AI assistants to tools and data. A service exposes an MCP server, and any MCP client, such as Claude or Cursor, can call its tools.