MCP integration

This section describes how to control OpenText Core Performance Engineering through external AI tools using the Model Control Protocol (MCP).

About MCP

The Model Context Protocol (MCP) is an open standard that enables AI assistants to securely connect with external tools and data sources. When you make a natural language request, such as "Give me the summary report for Run 2384", the MCP client translates this into a standardized protocol format and sends it to the MCP server.

The server then executes the appropriate operations using the OpenText Core Performance Engineering Public API, retrieves the data, and returns the response. This seamless flow transforms your conversational requests into actual operations, eliminating the need to manually navigate the UI or write API scripts.

Getting Started

To use the MCP server with OpenText Core Performance Engineering:

Download the MCP server image from Tools and integrations > MCP Server.
Run the image using the Docker pull command specified in the repository.
Follow the configuration instructions provided in the repository to set up the server and attach it to your MCP client. For example, Claude Desktop or Cursor.

MCP server Capabilities

The MCP server supports over 30 tools that map directly to the platform’s public API operations. These tools enable comprehensive test management and monitoring through natural language.

The supported tools includes the following core operations:

Area	Operation
Project management	Get all projects, get project details, get project users, get project scripts, get project load generators, get project load tests
Test configuration	Get load test details, get test scripts, get test settings, get test load generators, get test locations, get test transactions, get test schedules, estimate test cost
Test execution	Start new test run, update test run status (stop, pause, resume)
Test creation	Create new load test, add scripts to load test
Results & monitoring	Get test runs, get active test runs, get test run status, get test run results, get test run transactions, get runtime TRT summary, get test run load generators, get test run IPs, download test run report
User & asset management	Get users, get user projects, get script information, get streaming agents

Each tool corresponds to a single public API operation, making it easy to chain operations together for complex workflows.

Usage Examples

Example 1: Quick Status Check

Scenario: You hear from your team member that he executed an important test recently and want to see the results.

You ask Claude: "I think Raju from my team executed a test in the last few days, can you check and report the overall results?"

Behind the scenes:

Claude retrieves all users and finds Raju Chemla.
Gets Raju’s associated projects.
Checks for recent test runs across those projects.
Retrieves detailed results including metrics, transactions, and any issues.
Presents a comprehensive report with status, performance metrics, transaction success rates, and recommendations.

Example 2: Automated Test Creation and Execution

Scenario: You want to create and run a regression test but don’t remember the exact project name or script details.

You ask Claude: "I need to create a test on a project related to 'regression', take the most popular DevWeb script I have, add it to the test, run it, and share a high-level report."

Behind the scenes:

Claude searches all projects and identifies the "Regression 25.4" project.
Retrieves all scripts and identifies "DevWeb Encryption Test" as a popular DevWeb script.
Creates a new load test in the regression project.
Adds the DevWeb script to the test with appropriate configuration.
Starts the test run.
Monitors the execution.
Retrieves and presents results including throughput, response times, transaction success rates, and VUser metrics.

This natural language approach eliminates the need to remember specific IDs, navigate multiple UI screens, or write API scripts. Just describe what you need, and Claude handles the technical details.