Chapter 1. MCP server overview (Technology Preview)


The MCP server is a Model Context Protocol (MCP) server that provides access to the Red Hat Advanced Cluster Management for Kubernetes search database, which stores information about Kubernetes resources across managed clusters.

You must have Red Hat Advanced Cluster Management installed to use the MCP server. See Get started with Red Hat Advanced Cluster Management.

Use the MCP server to query cluster resources from AI assistants and other MCP clients. The server shows tools that search the same data that powers the Red Hat Advanced Cluster Management search service.

The MCP server provides output controls and filters for your artificial intelligence agents to use.

  • Available tools: The MCP server provides the find_resources tool for advanced Kubernetes resource search across all managed clusters.
  • Basic filters: kind, name, namespace, cluster, status
  • Advanced filters: labelSelector, clusterSelector, textSearch, ageNewerThan, ageOlderThan
  • Output control: outputMode (list, count, summary, or health), groupBy, sortBy, sortOrder, limit, countOnly

Learn how to install the MCP server and configure OpenShift Lightspeed or Claude Code to use it:

1.1. Install the MCP server (Technology Preview)

Install the MCP server from the published Helm repository on your Red Hat Advanced Cluster Management hub cluster. The Helm chart automatically discovers your Red Hat Advanced Cluster Management MultiClusterHub installation, extracts database credentials, and deploys the MCP server with the required configuration.

Required access: Cluster administrator

Prerequisites

Procedure

  1. Add the MCP server Helm repository. Run the following command:

    helm repo add acm-search https://raw.githubusercontent.com/stolostron/search-mcp-server/main/charts
    helm repo update
  2. Install the MCP server with auto-discovery. Run the following command:

    helm install acm-mcp-server acm-search/acm-mcp-server \
      --create-namespace \
      --namespace acm-search

    The Helm chart automatically completes the following tasks:

    • Discovers your Red Hat Advanced Cluster Management MultiClusterHub installation
    • Builds the complete database connection URL
    • Deploys the MCP server with the required configuration
  3. Verify that the MCP server is running in the acm-search namespace. Run the following command:

    oc get pods -n acm-search

    After you install the MCP server, you can connect your AI assistants or MCP clients.

Configure OpenShift Lightspeed to use the MCP server so that you can query Red Hat Advanced Cluster Management search data from the Lightspeed assistant.

Required access: Cluster administrator

Prerequisites

  • The MCP server is installed on your Red Hat Advanced Cluster Management hub cluster. See Installing the MCP server.
  • OpenShift Lightspeed is installed and configured on the Red Hat Advanced Cluster Management hub cluster where you want to add the MCP server. For more information, see Red Hat OpenShift Lightspeed installation documentation.

Important:

  • If you already have MCP servers configured in your OLSConfig custom resource, add the MCP server to the existing list rather than replacing it.
  • Use the exact configuration that is shown in this topic for proper functionality.
  • The acm-search, which is the default namespace for the Helm chart, is used as the service URL.

Procedure

  1. Edit the OLSConfig custom resource to add the MCP server. Run the following command:

    oc edit olsconfig cluster
    • If you do not have MCP servers configured in your OLSConfig custom resource, add featureGates and mcpServers to the following configuration:

      spec:
        featureGates:
          - MCPServer
        mcpServers:
          - headers:
              - name: Authorization
                valueFrom:
                  type: kubernetes
            name: acm
            timeout: 30
            url: 'http://acm-mcp-server.acm-search.svc.cluster.local/mcp'
    • If your OLSConfig custom resource already has MCP servers configured, add the MCP server to the existing mcpServers list. Your configuration might resemble the following example:
    mcpServers:
      - name: existing-server
        # ... existing server config ...
      - name: acm
        streamableHTTP:
          enableSSE: false
          headers:
            kubernetes-authorization: kubernetes
          sseReadTimeout: 10
          timeout: 30
          url: 'http://acm-mcp-server.acm-search.svc.cluster.local/mcp'

Additional resources

Configure and run Claude Code on your machine to connect to the MCP server that is installed on your Red Hat Advanced Cluster Management hub cluster. After you configure Claude Code, you can query Red Hat Advanced Cluster Management search data.

Prerequisites

Procedure

  1. Log in to the Red Hat Advanced Cluster Management hub cluster by using the oc command. Note: If your cluster uses a self-signed certificate, start Claude by setting the NODE_TLS_REJECT_UNAUTHORIZED environment variable to 0 in your command-line environment:

    export NODE_TLS_REJECT_UNAUTHORIZED=0
  2. Retrieve the route that was created by the MCP server installation in the acm-search namespace:

    export ROUTE=$(oc get route acm-mcp-server -n acm-search -o jsonpath='{.spec.host}')
  3. Run the following command to add the MCP server to Claude Code:

    claude mcp add --transport http -s project acm-search https://${ROUTE}/mcp --header "Authorization: Bearer $(oc whoami -t)"
  4. Log in to Claude Code, connect to the MCP server named acm-search, use the tools.
Red Hat logoGithubRedditYoutubeTwitter

Learn

Try, buy, & sell

Communities

About Red Hat

We deliver hardened solutions that make it easier for enterprises to work across platforms and environments, from the core datacenter to the network edge.

Making open source more inclusive

Red Hat is committed to replacing problematic language in our code, documentation, and web properties. For more details, see the Red Hat Blog.

About Red Hat Documentation

Legal Notice

Theme

© 2026 Red Hat
Back to top