The full upstream README, mirrored here for reference. Install config, tool schemas, adoption signals, and an original overview live on the SmartBear MCP listing page.
A Model Context Protocol (MCP) server which provides AI assistants with seamless access to SmartBear's suite of testing and monitoring tools, including BearQ, BugSnag, Reflect, Swagger, PactFlow, Pact Broker, QMetry, QTM4J, Zephyr and Collaborator.
The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect to external data sources and tools. This server exposes SmartBear's APIs through natural language interfaces, allowing you to query your testing data, analyze performance metrics, and manage test automation directly from your AI workflow.
See individual guides for suggested prompts and supported tools and resources:
For BugSnag, Swagger, and Zephyr, SmartBear hosts Remote MCP Servers that you can connect to directly from your MCP client via a URL — no installation, Node.js, or API tokens required. Authentication is handled through an OAuth browser flow.
| Product | Server URL |
|---|---|
| Swagger | https://swagger.mcp.smartbear.com/mcp |
| BugSnag | https://bugsnag.mcp.smartbear.com/mcp |
| Zephyr | https://zephyr.mcp.smartbear.com/mcp |
See the Remote MCP Servers guide for per-client setup instructions. You can connect to multiple remote servers at the same time.
Need BearQ, Reflect, QMetry, QTM4J, PactFlow, Collaborator, or Functional Testing? These products are only available via the local npm package below, which bundles all products into a single MCP server.
For all products — or if you prefer running the server locally — the MCP server is distributed as an npm package @smartbear/mcp, making it easy to integrate into your development workflow.
The server is started with the API key or auth token that you use with your SmartBear product(s). They are optional and can be removed from your configuration if you aren't using the product. For BugSnag, if you provide a project API key it will narrow down all searches to a single project in your BugSnag dashboard. Leave this field blank if you wish to interact across multiple projects at a time.
For the quickest setup, use the "MCP: Add server…" command in the Command Palette to add the @smartbear/mcp npm package.
Alternatively, you can use npx (or globally install) the @smartbear/mcp package to run the server and add the following to your .vscode/mcp.json file:
Add the following configuration to your claude_desktop_config.json to launch the MCP server via npx:
| Variable | Default | Purpose |
|---|---|---|
BASE_URL | (derived from request) | The server's own public URL, e.g. https://mcp.example.com. Set this on any deployment behind a proxy or load balancer — otherwise the address advertised for OAuth discovery is derived from the client-supplied Host header. |
TRUST_PROXY | false | Set to true only when a proxy in front of the server sets X-Forwarded-Host and strips any client-supplied value. |
For detailed introduction, examples, and advanced configuration visit our 📖 Full Documentation
For developers who want to contribute to the SmartBear MCP server, please see the CONTRIBUTING.md guide.
This MCP server is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License. For more details, please see the LICENSE file in the project repository.
SmartBear MCP Server - Bringing the power of SmartBear's testing and monitoring ecosystem to your AI-powered development workflow.