Axinova MCP Server is a Go-based implementation that exposes 40 DevOps tools across six services via the MCP protocol for native MCP clients (Claude Desktop, Claude Code, GitHub Copilot) and an HTTP API for non-MCP clients. The server runs tooling for Portainer (Docker management), Grafana (dashboards and data sources), Prometheus (metrics), SilverBullet (note-taking), Vikunja (task management), and Registry UI (Docker registry). Install the prebuilt binary or build from source, configure environment variables for access to your internal Portainer, Grafana, and TLS settings, then run the server either in development or production mode. You can interact with the tools either through the MCP-native transport (stdio) or via the HTTP JSON-RPC API depending on client support.

Prerequisites:

• Go 1.22+ (for building from source)
• Access to Axinova internal tools (ax-tools, ax-sas-tools)
• API tokens for each service (Portainer, Grafana, etc.)

Installation steps (recommended for development):

1. Clone the repository and install dependencies git clone https://github.com/axinova-ai/axinova-mcp-server-go.git cd axinova-mcp-server-go go mod tidy

2. Build the server make build

3. Run the server (development) make run

Production installation (prebuilt binary):

1. Download the latest binary from the release page

   Example (macOS)

   curl -L https://github.com/axinova-ai/axinova-mcp-server-go/releases/latest/download/axinova-mcp-server-macos -o /usr/local/bin/axinova-mcp-server chmod +x /usr/local/bin/axinova-mcp-server

2. Create a configuration file or export environment variables and run the binary

   Example with env vars:

   export APP_PORTAINER__URL=https://portainer.axinova-internal.xyz export APP_PORTAINER__TOKEN=your-token export APP_GRAFANA__URL=https://grafana.axinova-internal.xyz export APP_GRAFANA__TOKEN=your-token export APP_TLS__SKIP_VERIFY=true /usr/local/bin/axinova-mcp-server

3. If using Docker, build and run via docker-compose as described in the repository docs.

Tips and common considerations:

• Ensure all required tokens and URLs are correct and accessible from the server environment.
• If using TLS skip verification, only enable in trusted internal networks.
• The Production Server base URL is https://mcp.axinova-ai.com; configure clients accordingly.
• Documentation is available in docs/ for API reference, LLM integration, and tool catalog. Use Koanf-based env var support for flexible configuration.
• For MCP-native clients, prefer the stdio transport for best performance; use HTTP API for non-MCP clients.
• When updating dependencies or rebuilding, run go mod tidy to keep modules in sync.
• If you encounter port conflicts, adjust APP_PORTAINER__URL or the server port in your environment.