home / mcp / devops practices mcp server

DevOps Practices MCP Server

Provides centralized DevOps practices and templates for infrastructure projects through an MCP server.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "ai-4-devops-devops-practices": {
      "command": "python",
      "args": [
        "/home/<username>/.mcp-servers/devops-practices/mcp-server.py"
      ]
    }
  }
}

You deploy and interact with the DevOps Practices MCP Server to access shared infrastructure standards and templates. This server centralizes practices, templates, and tooling so your projects stay consistent, auditable, and easy to evolve.

How to use

You connect to the MCP server from an MCP client or Claude to query available practices, fetch practice content by name, list templates, retrieve templates, and render templates with specific values. Use these capabilities to discover standard procedures, obtain ready-to-use templates, and generate project documents with automatic variable substitution.

How to install

Follow these concrete steps to install and run the MCP server locally before connecting your client.

# 1. Choose a working directory and clone the MCP server repo
git clone <repo-url> ~/.mcp-servers/devops-practices
cd ~/.mcp-servers/devops-practices

# 2. Install dependencies (prefer uv for performance)
# Install uv if not already installed
curl -LsSf https://astral.sh/uv/install.sh | sh

# Install Python dependencies
uv pip install -r requirements.txt

# Alternative (traditional pip):
pip install -r requirements.txt

# 3. Configure the MCP server in Claude
# Create or edit Claude configuration to point to the MCP server (local stdio method shown below)

The recommended MCP server configuration uses a local, stdio-based setup. You run the server with Python and point Claude to the local script path as shown in the example configuration.

{
  "mcpServers": {
    "devops-practices": {
      "command": "python",
      "args": ["/home/<username>/.mcp-servers/devops-practices/mcp-server.py"],
      "env": {}
    }
  }
}

Additional sections

Configuration and startup details provide the practical path to getting results. The MCP server ships a Python implementation with a script you run locally. The server is designed to be queried by your MCP client to retrieve and render shared practices and templates.

Health checks and CI tooling help ensure reliability. You can validate the MCP server locally using the included health-check.sh script to verify structure, Python environment, and loading capabilities before integrating with Claude or your CI pipelines.

If you need to contribute, follow the governance workflow: create feature branches from develop, run the health check, update practice/template files, and create a merge request after passing CI. For releases, create release branches, update version numbers, and tag releases accordingly.

Available tools

list_practices

List all available practices stored in the MCP server.

get_practice

Fetch the content of a specific practice by name.

list_templates

List all available templates stored in the MCP server.

get_template

Fetch the content of a specific template by name.

render_template

Render a template with provided variables, returning the substituted content.