Testkube MCP

What testkube-mcp MCP server does

The testkube-mcp MCP server gives AI assistants access to Testkube’s testing platform. Testkube defines and runs automated tests in Kubernetes, using existing tools and scripts for API, end-to-end, performance, security, infrastructure, and other testing scenarios. Through MCP, an assistant can work with Testkube workflows, initiate or inspect executions, and access artifacts and collected results.

This is intended for teams that already use Testkube or want an assistant-facing interface for Kubernetes-based test operations. It does not replace the Testkube agent or testing infrastructure; it provides another way to interact with those capabilities.

How it works

Testkube runs tests inside Kubernetes infrastructure. Workflows describe how tests are organized and executed, while executions represent individual runs. Test results, logs, artifacts, and resource metrics are aggregated by Testkube for troubleshooting and reporting.

The testkube-mcp MCP server makes supported Testkube operations available to an AI assistant through the Model Context Protocol. Test runs can also be triggered through other Testkube mechanisms, including manual actions, schedules, CI/CD or GitOps pipelines, Kubernetes events, and the REST API. MCP is one interaction path among these options.

The repository describes two Testkube deployment models. The open-source agent runs independently in a Kubernetes cluster without a control plane. A commercial control plane can connect agents across clusters, teams, and environments, with additional enterprise features.

Setup and configuration

Testkube’s open-source agent can be installed in a target Kubernetes cluster with Helm or the Testkube CLI. The project also provides a quickstart for setting up Testkube and running initial tests.

Choose the standalone agent when a self-managed or single-cluster deployment is appropriate. The commercial control plane is described for organizations coordinating agents across multiple clusters and environments.

Tools and capabilities

The testkube-mcp MCP server supports the following capability areas described by the project:

• Manage Testkube test workflows.
• Trigger and inspect test executions.
• Review test results and execution information.
• Work with generated test artifacts.
• Use aggregated logs and resource metrics for troubleshooting and reporting.

Consult the project’s MCP documentation for those implementation details before building an automated client around specific operations.

Limitations and notes

This listing does not establish compatibility with a particular desktop editor or coding assistant. The repository excerpt also does not document whether the MCP endpoint uses stdio, HTTP, or another transport, so client configuration should be verified against the current project documentation.

The server’s usefulness depends on access to a configured Testkube environment and, for the open-source deployment, a Kubernetes cluster running the Testkube agent. Commercial control-plane features are described separately from the open-source agent.