Lead Image © lassedesignen, 123RF.com

Lead Image © lassedesignen, 123RF.com

Integrating AI Systems

Common Language

Article from ADMIN 93/2026
By
The Model Context Protocol helps standardize and simplify communication between applications and large language models. We take a closer look at the protocol and offer some practical examples.

The Model Context Protocol (MCP) is an open standard developed to enable seamless integration between large language models (LLMs), applications (tools), and external data sources. Anthropic [1] first announced the protocol back in November 2024 and is therefore officially considered its main developer. At the time of writing this article, version 2025-06-18 of the protocol specification had just been released (version 2025-11-25 has since been released). Anthropic looked to establish the protocol as an open standard from the outset and is working closely with the community to push the protocol's development forward. Besides the GitHub repository [2], the project website has comprehensive documentation of the protocol specification [3].

Standardized Communication

Just as USB created a universal interface that lets any USB device connect to any USB port, eliminating the need for device-specific connectors, MCP creates a standardized way for AI models to connect to different data sources and applications. Before the introduction of USB, each device needed its own proprietary connector. Similarly, before MCP, developers had to create custom integrations for each combination of AI application and data source.

MCP establishes a standard that lets any compatible client work with any compatible server, which significantly reduces the complexity of integrating AI models, because MCP eliminates the need to develop individual interfaces for different applications. For example, natural language can communicate with a specific application and query data.

Imagine an LDAP server with a large user database. Users need an LDAP client to read the data from the server. On a terminal, this typically would be the ldapsearch tool.


...

Use one of the options below to read the full article

Buy this article as PDF

Download Article PDF now with Express Checkout
Price $2.95
(incl. VAT)

Buy ADMIN Magazine

SINGLE ISSUES
Print Issues
Digital Issues
SUBSCRIPTIONS
Print Subscriptions
Digital Subscriptions

Related content

  • Lead Image © smileus, 123RF.com
    Coordinating distributed systems with ZooKeeper
    Anyone who manages many clusters should be meticulous in ensuring that organized processes prevail in the distributed server zoo. We describe how Apache ZooKeeper fulfills this task.
    more »
  • Lead Image © Galina Peshkova, 123RF.com
    The advantages of configuration management tools
    Etcd, ZooKeeper, Consul, and similar programs are currently the subject of heated debate in the world of configuration management. We investigate the problems they seek to solve and promises they make.
    more »
  • Lead Image © tiero, 123RF.com
    OpenLDAP Workshop
    Centralized user management with LDAP or Active Directory is the standard today, although many prefer to manage user data manually rather than build this kind of infrastructure. In this article, we look at a better approach with OpenLDAP.
    more »
  • Photo by Conner Baker on Unsplash
    Operating large language models in-house
    An internal AI server is an interesting way to retain data sovereignty. We show you how to set up an in-house AI server on your hardware and use it in parallel with AI services such as ChatGPT in the cloud.
    more »
  • © Nikita Sobolkov, 123RF.com
    The System Security Services Daemon
    The current version of the System Security Services Daemon (SSSD) lets you evaluate sudo rules centrally on an LDAP server, even if the LDAP server is inaccessible.
    more »
comments powered by Disqus