MCP Server

The vREST NG MCP (Model Context Protocol) server lets you use vREST NG from your IDE or other MCP clients (e.g. Cursor). The MCP server is started and used via the vREST NG CLI: you run the mcp command from the command line, then connect your MCP client to it. Once the server is running, you can run tests, read specifications, manage test cases, and use other vREST tools directly from the client.

What you can do with the MCP server

• Run vREST NG test cases from the command line or from your IDE
• Read and work with API specifications
• Create, update, and manage test cases and test suites
• Use vREST tools (e.g. list test suites, create test cases) from an MCP-capable application

How to use the MCP server

1. Generate a CLI token
   Create a token from the vREST NG Dashboard: Configuration > CLI Token > Generate CLI Token.
   For details, see Generate CLI Token (opens new window).

2. Start the MCP server via CLI
   From a terminal, run the vREST NG CLI with the mcp command, passing your token (and optionally a port):

   vrest-ng-cli mcp --token="<YOUR_CLI_TOKEN>" [--port=3000]
   

   The server listens on the given port (default: 3000).

3. Configure your MCP client
   In your client (e.g. Cursor), add an MCP server that connects to the running vREST MCP server (HTTP, on the port you used). See your client’s documentation for how to add and configure MCP servers.

More information

• For full CLI options and examples, see the MCP Server (CLI) guide.
• For generating and using the CLI token, see CLI Token (opens new window).