MCP - Model Context Protocol for Joomla!
A Joomla plugin providing a task-based API to manage content and connect with AI, workflow automation, and internal tools.
MCP - Model Context Protocol for Joomla!
MCP is a lightweight, powerful system plugin for Joomla that provides a Model Context Protocol—a simple, task-based API for managing content. It acts as a streamlined bridge between your Joomla website and the modern world of AI, workflow automation, and internal tooling. Learn more at Model Context Protocol.
The Problem
Joomla's built-in REST API is comprehensive, but its complexity can be a hurdle for rapid integration. Modern AI agents and workflow automation platforms thrive on simple, predictable "tools" or endpoints to perform specific actions. They need a clear protocol to interact with external data sources without requiring complex, service-specific configurations.
The Solution: MCP (Model Context Protocol)
MCP establishes a simple protocol for AI models and services to interact with your Joomla content. It provides a flat, single-endpoint API where an action is specified by a task
parameter. This design makes it incredibly easy for any application to create, update, or retrieve articles and categories—providing and receiving context through a standardized protocol.
It authenticates using Joomla's native API Token system, ensuring that all operations are secure and respect the permission levels of the user associated with the token.
Key Features
- Simple Protocol: No complex RESTful routes to learn. Just one URL and a
task
parameter. - Secure: Leverages Joomla's core
Web Services - Authentication - Token
system. - Lightweight: A single plugin with no external dependencies.
- AI-Ready: Designed to serve as the perfect "tool" for AI agents and automation workflows to read from and write to your Joomla database.
- Essential Endpoints: Covers the most common content management tasks.
The AI & Workflow Automation Superpower
This is where MCP truly shines. It transforms your Joomla CMS from a siloed platform into a dynamic, integrated component of your automated workflows and AI-driven content pipelines.
For Workflow Automation (Make.com, n8n, Zapier, etc.)
Platforms like Make.com and n8n are built around connecting services through API calls. MCP provides the perfect endpoints for their generic "HTTP Request" modules.
**Example Use Case:**An RSS feed triggers a workflow in n8n. An AI node rewrites the content, and then an HTTP Request node uses the Model Context Protocol (
task=mcp.create_article
) to instantly publish the new article to your Joomla site.
For AI Agent Frameworks (crewAI, AutoGen, etc.)
AI agents need "tools" to interact with the real world. MCP's endpoints are the perfect building blocks for these tools, allowing agents to autonomously manage content.
**Example Use Case:**You task a crewAI agent with "Write a blog post about the latest AI trends and publish it to our website."
- The
ResearcherAgent
browses the web.- The
WriterAgent
composes the article.- The
PublisherAgent
is given apublish_to_joomla
tool, which uses the Model Context Protocol to call themcp.create_article
endpoint. The AI crew completes the entire task without human intervention.
For Internal Tooling & Scripting (Windmill, Superblocks, etc.)
Platforms like Windmill allow you to quickly build internal admin panels and run scripts. MCP provides a clean abstraction layer for interacting with Joomla.
**Example Use Case:**Your marketing team wants a simple dashboard in Windmill to quickly publish press releases. A developer creates a simple UI. The "Publish" button triggers a script that uses the Model Context Protocol to push the content live instantly.
Installation & Setup
- Download: Download the latest
plg_system_mcp_vX.X.X.zip
file from the Releases page. - Install: In your Joomla Administrator panel, go to
System
->Install
->Extensions
and upload the zip file. - Enable Plugin: Go to
System
->Manage
->Plugins
and search for "MCP". Enable the plugin. - Generate a User API Token:
- Go to
Users
->Manage
and select the user you want to grant API access to. This user's permissions will be respected. - Click the "Joomla API Token" tab.
- Click "Create a New Token" to generate an API key. Copy this key securely.
- Go to
API Usage
Note: This is Version 2.0 with consolidated API tasks and improved architecture.
- Endpoint:
https://www.yoursite.com/index.php
- Method:
POST
- Authentication Header:
X-Joomla-Token: YOUR_JOOMLA_API_TOKEN
- Query Parameter:
task=mcp.your_task
Available Tasks
Task | Description | Example JSON Body |
---|---|---|
mcp.info | Returns plugin info and authentication status. | null |
mcp.get_article | Retrieves a single article by ID. | {"article_id": 124} |
mcp.get_articles | Retrieves a list of all articles. | null or {"catid": 8, "state": 1, "limit": 10} |
mcp.get_categories | Retrieves a list of all content categories. | null |
mcp.get_tags | Retrieves a list of all tags. | null |
mcp.create_article | Creates a new article. | {"title": "My Title", "articletext": "Content", "catid": 2, "published": true} |
mcp.update_article | Updates article content and/or state (consolidated). | {"article_id": 124, "title": "Updated Title", "state": 0} or {"article_id": 126, "state": -2} |
Example: Full curl
Request
Here's how to create a new, published article in category ID 8
.
curl -X POST
-H "X-Joomla-Token: YOUR_JOOMLA_API_TOKEN"
-H "Content-Type: application/json"
-d '{"title": "New Article via MCP", "articletext": "This content was published by an automated workflow!", "catid": 8, "published": true}'
"https://www.yoursite.com/index.php?task=mcp.create_article"
Using MCP with Workflow Automation Tools
n8n Integration
- Create an HTTP Node:
- Add an HTTP Request node to your workflow.
- Set the Method to
POST
. - Set the URL to
https://<your_joomla_website_url>/index.php
. - Add a query parameter:
task=mcp.<your_task>
(e.g.,task=mcp.create_article
).
- Add Headers:
- Add a header:
X-Joomla-Token
with the value<your_joomla_api_token>
.
- Add a header:
- Add JSON Body:
- Add the JSON payload for the task you want to perform (e.g., creating or updating an article).
Make.com Integration
- Create a Scenario:
- Add an HTTP module to your scenario.
- Set the Method to
POST
. - Set the URL to
https://<your_joomla_website_url>/index.php
.
- Add Query Parameters:
- Add a query parameter:
task=mcp.<your_task>
(e.g.,task=mcp.get_article
).
- Add a query parameter:
- Add Headers:
- Add a header:
X-Joomla-Token
with the value<your_joomla_api_token>
.
- Add a header:
- Add JSON Body:
- Add the JSON payload for the task you want to perform.
Version 2.0 Changes
Implemented Improvements
- Task Consolidation:
- The
manage_article_state
andmove_article_to_trash
tasks have been consolidated into theupdate_article
task. You can now update content and state in a single call. - Use
{"article_id": 123, "state": -2}
to move to trash,{"article_id": 123, "state": 0}
to unpublish, etc.
- The
- New Features:
- Added
mcp.get_tags
task for retrieving all tags - Added
mcp.info
task for plugin information and authentication status - Enhanced filtering and pagination support for
get_articles
- Added
- Code Architecture:
- Completely refactored with organized handler classes
- Improved authentication system
- Better error handling and validation
- Streamlined JSON responses with timestamps
Related Servers
openapi-to-mcp
Expose API endpoints as strongly typed tools from an OpenAPI specification. Supports OpenAPI 2.0/3.0 in JSON or YAML format, from local or remote files.
Moralis
Interact with the Moralis Web3 API to access blockchain data and services.
Remote MCP Server (Authless)
An example of a remote MCP server without authentication, deployable on Cloudflare Workers.
ContextKeeper
Provides perfect memory for AI-assisted development by capturing project context snapshots, enabling natural language search, evolution tracking, and code intelligence.
Unified MCP Client Library
A TypeScript library for integrating MCP with tools like LangChain and Zod, providing helpers for schema conversion and event streaming.
Godot MCP
MCP server for interacting with the Godot game engine, providing tools for editing, running, debugging, and managing scenes in Godot projects.
Stata-MCP
Perform regression analysis using Stata with the help of an LLM. Requires a local Stata installation and an external LLM API key.
Tabby-MCP-Server
A Tabby plugin implementing an MCP server for AI-powered terminal control and automation.
AI Agent Timeline MCP Server
A timeline tool for AI agents to post their thoughts and progress while working.
MCP for Docs
Automatically downloads and converts documentation from various sources into organized markdown files.