AI Tip of the Week #6: Integrating Tricentis Tosca Cloud with MCP (Model Context Protocol)

Overview

Tricentis Tosca Cloud now supports the Model Context Protocol (MCP) integration, allowing QA engineers to use AI assistants (like Cursor AI or Visual Studio Code) to interact with Tosca Cloud via natural language. In practical terms, MCP lets you perform Tosca Cloud testing tasks directly from your development environment – no more switching between tools. You can automate test case creation, analyze failures, and even simulate APIs right from your IDE. Below, we explain what MCP enables, how to configure it securely, and the new workflows it unlocks.

What MCP Enables in Tosca Cloud

MCP is essentially a bridge between Tosca Cloud and your AI-powered development tools. Here’s what it makes possible for your testing workflow:

• AI-driven test design: Automate tedious test case creation by letting the AI assistant scaffold test cases for your application flows. This accelerates test design and ensures you cover key user scenarios.

• Smart failure analysis: Gain insights into failed test runs by asking the AI to diagnose errors and suggest improvements. This helps identify root causes faster and improve your tests iteratively.

• On-the-fly API simulation: Speed up development by generating API simulations on demand. You can simulate external services or endpoints within Tosca Cloud, so your testing isn’t blocked by missing components – all controlled via simple commands.

• Seamless in-IDE control: All these tasks are performed from within your IDE or code editor using natural language commands, meaning you don’t have to leave your development environment. This reduces context switching and streamlines your QA workflow.

Secure Setup: Configuring the MCP Integration

Before connecting your AI assistant to Tosca Cloud’s MCP server, make sure you have the following prerequisites ready:

• Tosca Cloud access: An active Tosca Cloud tenant (account) that you can sign in to.

• Node.js installed: MCP uses a Node.js tool under the hood, so install Node.js on your system if not already present.

• Supported AI assistant: Either Cursor AI or Visual Studio Code (with an AI chat extension) is required to serve as the front-end assistant.

• VS Code org setting: If you use VS Code, ensure your organization has enabled the MCP integration setting (chat.mcp.enabled) so the feature isn’t blocked.

Important: The MCP server communicates over port 56874. Keep this default port; changing it will cause connection errors.

Setup Steps:

Once prerequisites are set, follow these steps to connect your AI tool to the Tosca Cloud MCP server:

1. Open the MCP configuration in your AI tool:

   • In Cursor AI: Go to Cursor’s settings and find the MCP Tools section. Click “New MCP Server” to create a new mcp.json configuration file.

   • In Visual Studio Code: Open your VS Code settings (e.g. your settings.json file). Ensure "chat.mcp.enabled": true is set, then prepare to add a new entry under the "mcp.servers" configuration for Tosca Cloud.

2. Add the Tosca Cloud MCP server details: In the new config, point your AI assistant to the Tosca Cloud MCP endpoint and credentials:

   • Endpoint URL: Use the URL for your Tosca Cloud tenant’s MCP API – e.g. https://<your-tenant>.my.tricentis.com/default/_mcp/api/mcp (replace <your-tenant> with your tenant name). If your Tosca Cloud uses a specific workspace other than the default, put the workspace ID in place of "default" in the URL.

   • Port and host: Use port 56874 and host 127.0.0.1 exactly as given (this ensures the MCP connection runs through your local machine).

   • OAuth credentials: Include the static OAuth client info provided by Tricentis: set the client ID to "MCPServer" and the scope to "tta" – these should remain exactly as provided. (These values ensure a secure, authorized connection to Tosca Cloud’s APIs.)

3. Save and enable the connection: Save the configuration file, then activate the MCP server integration in your AI assistant. For Cursor, this means enabling the new ToscaCloudMcpServer entry you added (you should see it listed in Cursor’s MCP Tools). In VS Code, saving the settings with chat.mcp.enabled will start the MCP client automatically on next use.

4. Authenticate with Tosca Cloud: Upon enabling, you’ll be prompted to sign in to Tosca Cloud. Log in with your credentials and complete any authentication steps required. This OAuth login grants your local MCP client permission to access your Tosca Cloud tenant securely.

5. Verify the connection: Finally, confirm that the AI assistant is connected to Tosca Cloud. In Cursor, look for a green status indicator and a list of available Tosca Cloud tools under the MCP server configuration – this means the connection is live. In VS Code, you can check that no error messages appear and that Tosca Cloud commands are available in your AI chat interface. You are now ready to use natural language to control Tosca Cloud! 🎉

(If you encounter issues, refer to Tricentis’s troubleshooting tips. For example, if you get a port error or the VS Code setting is blocked by admin policy, you may need to restart your AI assistant or have your admin enable MCP for your organization.)

New Workflows Unlocked with MCP

Once MCP is active, you can integrate Tosca Cloud capabilities directly into your daily workflow using Cursor or VS Code. Here are some of the powerful things you can do without leaving your editor:

• Generate test cases with AI: Ask your assistant to create new test cases or modules by describing a scenario. For example, you could prompt it to “Scaffold a test case for the user registration flow”, and Tosca Cloud will automatically generate the test case structure. This accelerates test design and ensures consistency in new tests.

• Analyze and debug test runs: Leverage AI to interpret test results. You can request something like “Show me the errors from the last test run” to get an analysis of failed steps or ask “How can I improve this test case?” for suggestions. The MCP integration will retrieve failure details and insights, helping you pinpoint issues quickly.

• Maintain and refactor tests: Easily perform maintenance tasks on your Tosca tests via chat commands. For instance, you can rename test steps to follow naming standards, clean up test step labels, or search for specific test artifacts (like all tests related to “registration”) just by issuing a request. This keeps your test repository organized with minimal manual effort.

• Simulate APIs on the fly: With MCP, you can create and deploy API simulations directly through AI instructions. If your test needs a backend service that isn’t available, simply tell the assistant to simulate that API. For example, “Create a simulated login API and deploy it” will create a stub service in Tosca’s API Simulation tool for you. This is extremely useful for testing in isolation and speeding up development cycles.

• Manage playlists and runs: Orchestrating test suites becomes conversational. You can tell the AI to create a new playlist (a set of test cases) for a specific purpose (e.g. regression tests) and then trigger its execution – all via MCP. After a run, you could ask “Get the latest playlist run results” to retrieve the execution log or outcomes immediately. This makes it effortless to run and monitor large groups of tests.

By configuring MCP and using these workflows, QA engineers can streamline test automation significantly. Routine tasks that once took multiple steps in the Tosca UI can now be accomplished by simply asking your AI assistant, right from VS Code or Cursor. This integration keeps you in the zone – write code, generate and run tests, analyze failures, and simulate needed services, all in one place without switching contexts.

Give MCP a try in your Tosca Cloud environment this week. With a secure setup in place, you’ll be able to harness the power of AI for testing and boost your productivity in a truly hands-on-keyboard fashion. Happy testing! 🚀

Visit this link to learn more