BrowserStack MCP Server
BrowserStack mobile test platform integration for AI-driven debugging and execution.
Data last scanned today · Reviewed today
Overview
The BrowserStack MCP Server provides a direct interface for interacting with BrowserStack’s cloud-based testing infrastructure from within your AI assistant or IDE. It allows users to manage test cases, trigger manual or automated mobile tests on real devices, and debug application crashes using natural language commands. By keeping testing workflows inside the developer environment, it aims to minimize context switching during the software development lifecycle. Designed for integration into any MCP-compatible client, the server utilizes a stateless architecture over Streamable HTTP to maintain connectivity. It supports complex testing operations across multiple operating systems, enabling developers to perform tasks like opening apps on specific mobile devices and analyzing crash reports without leaving their coding workflow.
Our verdict
This server carries a high-risk rating of 45/100, primarily because the authentication methods and read-only modes have not yet been editorially reviewed or independently verified. Given that it interacts with external testing platforms, users should exercise caution regarding the credentials used to link their accounts. It is suitable for developers looking to automate mobile testing workflows, but should be used in controlled environments until further security analysis is performed.
- Supports active development with recent commits
- Maintained by a large community of 21 contributors
- Streamable HTTP architecture simplifies client connectivity
- Provides a manageable attack surface with 13 direct dependencies
- Security mechanisms remain unevaluated by external reviewers
- Authentication and read-only access modes are unverified
- Not developed or maintained by the official BrowserStack organization
Setup
{
"mcpServers": {
"browserstack-mcp-server": {
"args": [
"-y",
"@browserstack/mcp-server"
],
"command": "npx"
}
}
}Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.
- No read-only mode — running this grants full read/write access.
- No built-in authentication — anyone with access to this config can invoke every tool it exposes.
Tools
| Tool | Description | Risk |
|---|---|---|
| createProjectOrFolder | Create a Test Management project and/or folders to organize test cases. | medium |
| createTestCase | Add a manual test case under a specific project/folder. | medium |
| updateTestCase | Update an existing test case. | medium |
| listTestCases | List test cases for a project, optionally scoped to a folder. | low |
| listFolders | List folders in a Test Management project. | low |
| listTestCaseTemplates | List test-case templates with their numeric template_id. | low |
| createTestRun | Create a test run (suite) for selected test cases in a project. | medium |
| listTestRuns | List test runs for a project (filter by dates, assignee, state). | low |
| updateTestRun | Update a test run's name/state and/or add test cases to it. | medium |
| addTestResult | Add a manual execution result (passed/failed/blocked/skipped) for a test case within a run. | medium |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›What are the environment requirements for running the BrowserStack MCP server?
You must have Node.js version 18.0 or higher installed, with version 22.15.0 (LTS) being the recommended version.
›Do I need to configure settings to connect the server to my client?
No, the hosted Remote MCP server is stateless over Streamable HTTP, allowing compatible clients like Cursor, Claude, and VS Code to connect without configuration changes.
›Does the server automatically apply fixes to my code?
No, when fetching root cause analysis or suggested fixes for test failures, the server only proposes changes. You must provide explicit approval before your assistant applies any edits to your files.
›Can I test web applications that are running on my local machine?
Yes, you can test websites hosted locally, such as those on localhost, without needing to deploy them to a remote server.
›Which automated testing frameworks are supported for running tests on the BrowserStack platform?
You can run tests written in various frameworks, including Playwright, Selenium, and Jest, using the BrowserStack test platform.
Alternatives
Changelog
Badge
Maintain this server? Add the live badge to your README.