Run web automation tests in BrowserStack Automate with the MCP server

Use natural language prompts in your AI assistant to instantly set up BrowserStack SDK, run web automation tests, and fetch screenshots from Automate sessions on real browsers and devices.

Use the BrowserStack MCP server to set up the BrowserStack SDK, run web automation tests, and fetch screenshots from Automate sessions directly from your coding assistant (such as GitHub Copilot, Claude, or Cursor). Skip manual configuration and dashboards—start your web tests or retrieve screenshots with simple prompts.

To get started with BrowserStack MCP server, see the get started with BrowserStack MCP documentation.

Streamline web automation testing from your IDE

Use the following tools to set up the BrowserStack SDK, run web automation tests, and fetch screenshots from BrowserStack Automate sessions, all through natural language prompts:

Run web automation tests (setupBrowserStackAutomateTests)

Use the setupBrowserStackAutomateTests tool to integrate the BrowserStack SDK and run your web automation tests on real browsers in the cloud. This is the primary tool for executing your test suites with support for various frameworks and programming languages.

Example: Using MCP server with Automate using GitHub Copilot agent on VS Code

Fetch screenshots (fetchAutomationScreenshots)

Use the fetchAutomationScreenshots tool to retrieve screenshots from your Automate test sessions. This tool helps you access and analyze visual test results from your automation runs.

Tool reference

Use the setupBrowserStackAutomateTests tool to integrate the BrowserStack SDK and run web tests, and the fetchAutomationScreenshots tool to fetch screenshots from Automate/App Automate sessions.

Common workflows

The following workflows are supported on the BrowserStack MCP server to accomplish complex testing scenarios. Each workflow shows the sequence of tools and automated actions that streamline common development tasks:

Rapid SDK setup and integration

Run your automated web testing project on BrowserStack platform by seamlessly integrating BrowserStack SDK, configuring devices, and enabling visual testing (Percy) in one go.
Tools invoked: setupBrowserStackAppAutomateTests, runAppTestsOnBrowserStack.

Example prompts:

• Set up Automate SDK for my Selenium (Python) project, configure it to run on Pixel 7 and iPhone 15, and enable Percy for visual testing.

• Set up BrowserStack SDK for my Appium project, run these automated tests on BrowserStack using MCP server.
  Project: [Insert GitHub URL]
  Framework: Selenium
  Language: Java
  Target Devices: Samsung Galaxy S23, iPhone 14 Pro
  Enable Percy: true (optional)
  

Automated actions (What the MCP server does):

1. Detects your project’s framework and language (e.g., Selenium + Java).

2. Installs the necessary BrowserStack SDK and dependencies into your project.

3. Configures the browserstack.yml file with your requested devices (e.g., Pixel 7, iPhone 15).

4. Enables Percy for visual testing if requested.

Run tests and debug failures

Streamline the testing process by automating execution, detecting failures, and leveraging AI for Root Cause Analysis (RCA) and resolution.
Tools invoked: runAppTestsOnBrowserStack, getBuildId, listTestIds, fetchRCA

Example prompts:

• “Run my Selenium tests on BrowserStack. Once finished, list any failed test IDs.”

• “List failed test IDs for build . Fetch the RCA for each failure and suggest code fixes to resolve them."

Automated actions (What the MCP server does):

1. Executes your test suite on the BrowserStack Automate cloud.

2. Retrieves the specific Build ID and filters the list for status=failed.

3. Fetches RCA logs to diagnose if the failure is due to an automation issue, a product bug, or an environment issue.

4. Suggests code fixes based on AI analysis of the logs (via Copilot).

Need help?

• FAQs for common questions.
• Troubleshoot for common issues.