This repo contains code for a lightweight client to interact with GLPI and a MCP to consume it from any LLM

Requirements

• Available glpi instance
• valid user and API token
• prettytable python library

Installation

pip3 install glpic

Using client

Store your creds in any env file such as glpic.env.sample and set data accordingly You can then use the following commands

• List computers

glpic list computers

• Get information on a specific computer

glpic info computer $computer

• List reservations

glpic list reservations

• Get information on a specific reservation

glpic info reservation $reservation

• Update a given reservations

glpic update reservation $reservation -P end=20240601

• Update all reservations

glpic update reservations

Using MCP

The server is started and configured differently depending on what transport you want to use

For STDIO:

In VSCode for example:

   "mcp": {
        "servers": {
            "glpi": {
                "command": "python3",
                "args": ["/path/to/glpic/src/glpic/mcp_server.py", "--stdio"],
                "env": {
                    "GLPI_URL": "https://server/apirest.php",
                    "GLPI_USER": "myuser",
                    "GLPI_TOKEN": "mytoken"
                }
            }
        }
    }

For Streamable HTTP:

Start the server in a terminal:

glpimcp

Configure the server in your client:

    "rhsupportcli": {
      "transport": "streamable-http",
      "url": "http://your_server:8000"
      "headers": {
        "GLPI_URL": "https://server/apirest.php",
        "GLPI_USER": "myuser",
        "GLPI_TOKEN": "mytoken"
      }
    }