io.github.kubeshop/testkube-mcp
MCP server for Testkube - Manage test workflows, executions, and artifacts via AI assistants
Installation
Ajoutez ceci à la configuration MCP de votre client (Claude Desktop : claude_desktop_config.json ; Cursor : .cursor/mcp.json ; VS Code et les autres utilisent la même forme « mcpServers »).
claude mcp add --transport http testkube-mcp 'https://api.testkube.io/organizations/{organization_id}/environments/{environment_id}/mcp'
{
"mcpServers": {
"testkube-mcp": {
"type": "http",
"url": "https://api.testkube.io/organizations/{organization_id}/environments/{environment_id}/mcp"
}
}
}
docker.io/kubeshop/mcp-server:2.15.0claude mcp add testkube-mcp -e 'TK_ACCESS_TOKEN=<your tk_access_token>' -e 'TK_ORG_ID=<TK_ORG_ID>' -e 'TK_ENV_ID=<TK_ENV_ID>' -e 'TK_CONTROL_PLANE_URL=<TK_CONTROL_PLANE_URL>' -e 'TK_DASHBOARD_URL=<TK_DASHBOARD_URL>' -e 'TK_DEBUG=<TK_DEBUG>' -e 'TK_MCP_SHTTP_AUTH_TOKEN=<your tk_mcp_shttp_auth_token>' -- docker run -i --rm docker.io/kubeshop/mcp-server:2.15.0
{
"mcpServers": {
"testkube-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"docker.io/kubeshop/mcp-server:2.15.0"
],
"env": {
"TK_ACCESS_TOKEN": "<your tk_access_token>",
"TK_ORG_ID": "<TK_ORG_ID>",
"TK_ENV_ID": "<TK_ENV_ID>",
"TK_CONTROL_PLANE_URL": "<TK_CONTROL_PLANE_URL>",
"TK_DASHBOARD_URL": "<TK_DASHBOARD_URL>",
"TK_DEBUG": "<TK_DEBUG>",
"TK_MCP_SHTTP_AUTH_TOKEN": "<your tk_mcp_shttp_auth_token>"
}
}
}
}
Variables d'environnement
TK_ACCESS_TOKENobligatoire secret Testkube access token for authenticationTK_ORG_IDobligatoire Testkube organization IDTK_ENV_IDobligatoire Testkube environment IDTK_CONTROL_PLANE_URLTestkube Control Plane URL (default: https://api.testkube.io)TK_DASHBOARD_URLTestkube Dashboard URLTK_DEBUGEnable debug mode (default: false)TK_MCP_SHTTP_AUTH_TOKENsecret Separate inbound bearer token, required only when explicitly using --transport=shttp; not needed for default stdio
Documentation
Tirée du README du projet sur GitHub, actualisée chaque semaine. Les liens ouvrent GitHub.
The Open Testing Platform for AI-driven Engineering Teams
Testkube provides a single platform for defining, running and analyzing automated tests, using your existing testing tools/scripts, running in your Kubernetes infrastructure.
Trusted by Engineering Teams at CoreWeave, NVidia, Siemens, T-Mobile, Harvard, SwissCom, and many more..
Website | Docs | Changelog | Blog | Slack | LinkedIn | X
Why Testkube?
- Run any Tests : Execute any tests/tools/scripts at scale; API, E2E, Performance, Security, Infrastructure, etc. - Examples & Guides.
- Trigger Tests whenever needed: Trigger tests manually, on schedules, from CI/CD/GitOps pipelines, on Kubernetes Events, via the REST API, through MCP, etc. - Read More.
- See Everything: All test results, artifacts, logs and resource-metrics are aggregated for centralized troubleshooting and reporting - Read More.
- Integrate Natively: Testkube integrates with existing tools and infrastructure using Webhooks, the Testkube REST API or the MCP Server - see Integration Examples.
- Testkube AI: Use the Testkube MCP Server or native AI Agents for troubleshooting, analysis, remediation, etc - Read More
- Enterprise Ready: SSO/SCIM, RBAC, Teams, Resource-Groups, Audit-logs, etc. - Read More.
See it in action: Open the interactive TestWorkflows showcase to see how a workflow builds up from a single test to a fully orchestrated pipeline.
Two ways to run Testkube
Open Source Agent - this repo.
MIT-licensed, runs standalone in your Kubernetes cluster, no control plane required. Great for single-cluster setups, self-managed environments, and evaluating Testkube.
- The Helm or CLI Installation will make it easy to deploy the agent into your target cluster.
- The Quickstart is the easiest way to set up Testkube and run your first tests.
Check out the Testkube Open Source Overview to learn more about the open source deployment architecture.
Commercial Control Plane
The control plane connects every Testkube agent across clusters, teams, and environments into a single dashboard:
- One control plane, unlimited clusters - orchestrate and analyze tests across clusters and regions
- Testkube AI - workflow generation, failure investigation, remediation PRs
- Enterprise-grade - SSO/SCIM, RBAC, audit logs, SLA-backed support
Check out the Installation Overview to learn more about different ways to deploy and run the Testkube Control Plane.
The online Trial is the easiest way to try the commercial Testkube offering - Get Started
Marketplace
The Testkube Marketplace provides an open and ready-to-use catalog of Testkube Workflows for Infrastructure Testing - Read More.
Documentation
Extensive documentation is available at docs.testkube.io.
Contributing
Check out our Contributors Guide and the Agent Architecture to find your way around our codebase and process.
If you want to contribute code, this reading order works well:
- CONTRIBUTING.md - contribution workflow, coding standards, and PR process
- DEVELOPMENT.md - local setup with Tilt and day-to-day development loop
- ARCHITECTURE.md - high-level system design and key code paths
Shout-out to our contributors 🎉 - you're great!
- ⭐️ @lreimer - K6 executor Gradle executor Maven executor
- ⭐️ @jdborneman-terminus - Ginkgo executor
- ️⭐️ @abhishek9686
- ⭐️ @ancosma
- ⭐️ @Stupremee
- ⭐️ @artem-zherdiev-ingio
- ⭐️ @chooco13 - Playwright executor
Go to contribution document to read more how can you help us 🔥
Feedback
Whether Testkube helps you or not, we would love to help and hear from you. Please join us on Slack to ask questions and let us know how we can make Testkube even better!
Questions fréquentes
Qu'est-ce que io.github.kubeshop/testkube-mcp ?
MCP server for Testkube - Manage test workflows, executions, and artifacts via AI assistants
Comment installer io.github.kubeshop/testkube-mcp ?
Il fonctionne comme serveur distant. Dans Claude Code : claude mcp add --transport http testkube-mcp 'https://api.testkube.io/organizations/{organization_id}/environments/{environment_id}/mcp'. Pour les autres clients, collez la configuration JSON de cette page dans leurs réglages MCP.
Faut-il une clé API pour io.github.kubeshop/testkube-mcp ?
Oui. Il attend TK_ACCESS_TOKEN, TK_ORG_ID, TK_ENV_ID, TK_MCP_SHTTP_AUTH_TOKEN, à renseigner dans sa configuration.
io.github.kubeshop/testkube-mcp est-il open source ?
Son code est public sur GitHub, mais le dépôt ne déclare pas de licence standard.
Qui maintient io.github.kubeshop/testkube-mcp ?
Il est publié par kubeshop. AI.tn le référence sans le maintenir ni le recommander.
Issu du registre MCP officiel. Être listé n'est pas une recommandation : vérifiez le code et les permissions d'un serveur avant de lui donner accès à vos données.