MCP RAN POC is an MCP server proof of concept that connects AI clients to PostgreSQL databases and Kubernetes clusters managed by Red Hat Advanced Cluster Management via the Topology Aware Lifecycle Manager. Designed for cloud engineers, telecom operators, and infrastructure administrators, it enables unified natural language control over multi-database environments and fleet lifecycle management. The system includes two dedicated servers: one for querying PostgreSQL instances and another for interacting with TALM. The PostgreSQL server provides read-only execution for alarms, resources, and clusters databases, restricting commands strictly to SELECT and WITH queries to prevent accidental modifications. The TALM server handles Kubernetes cluster remediations, cluster group upgrades, health checks, and policy evaluations. In addition to serving standalone MCP clients like Claude Desktop, the repository includes a Google ADK web agent interface that connects both servers simultaneously. This enables operators to analyze telemetry, inspect cluster topologies, and trigger lifecycle remediations through a single conversational workspace.
Category: Cloud & Infrastructure
Tags: cluster-management, database, kubernetes, SQL
To install and configure MCP RAN POC with Claude Desktop, follow these steps: 1. Clone the repository and install dependencies locally using uv sync. 2. Open your Claude Desktop configuration file (claude_desktop_config.json). 3. Add both the talm and postgres servers under the mcpServers object: json { "mcpServers": { "talm": { "command": "uv", "args": ["run", "python", "servers/talm.py"], "cwd": "/path/to/mcp-ran-poc", "env": { "KUBECONFIG": "/path/to/your/kubeconfig.yaml" } }, "postgres": { "command": "uv", "args": ["run", "python", "servers/ocloud-pg.py"], "cwd": "/path/to/mcp-ran-poc", "env": { "ALARMS_DB_HOST": "your-postgres-host", "ALARMS_DB_PORT": "5432", "ALARMS_DB_NAME": "alarms", "ALARMS_DB_USER": "alarms", "ALARMS_DB_PASSWORD": "your-password", "RESOURCES_DB_HOST": "your-postgres-host", "RESOURCES_DB_PORT": "5432", "RESOURCES_DB_NAME": "resources", "RESOURCES_DB_USER": "resources", "RESOURCES_DB_PASSWORD": "your-password", "CLUSTERS_DB_HOST": "your-postgres-host", "CLUSTERS_DB_PORT": "5432", "CLUSTERS_DB_NAME": "clusters", "CLUSTERS_DB_USER": "clusters", "CLUSTERS_DB_PASSWORD": "your-password" } } } } 4. Update file paths, database hostnames, and credentials, then restart Claude Desktop.
Part of MCP Servers
MCP RAN POC is a proof of concept Model Context Protocol system containing dual MCP servers for PostgreSQL database interaction and Kubernetes cluster management via Red Hat ACM Topology Aware Lifecycle Manager. It also provides a Google ADK web agent interface for natural language operations.
No, the repository explicitly specifies that it is a proof of concept and is not designed or certified for production use. It is intended for testing, development, and architectural validation.
MCP RAN POC supports Claude Desktop via local stdio process definitions and streamable HTTP transports. It also includes a custom Google ADK agent that runs a web-based client interface accessible in modern web browsers.
No, the PostgreSQL MCP server restricts query execution to read-only statements. Only SELECT and WITH queries are permitted through the execute_query tool to ensure safe database inspections.
Yes, the repository provides deployment manifests and make targets to build container images, push them to an image registry, and deploy the entire multi-container pod to OpenShift Container Platform using kustomize.