What is the Model Context Protocol?

Understanding the foundation of AI-powered content management

A Protocol for AI Integration

The Model Context Protocol (MCP) is a standardized framework that enables AI clients (like Claude, ChatGPT, or custom AI agents) to safely and efficiently interact with external systems and data sources.

Think of MCP as a universal translator between AI assistants and your business applications. It provides a consistent interface for AI models to discover capabilities, request actions, and access information—without needing custom integrations.

How AI Clients Use MCP

✓ Discovery: AI clients automatically discover available tools and resources through the MCP protocol
✓ Safe Execution: All actions are validated and authenticated before execution
✓ Context Awareness: AI maintains full context about your content structure and relationships
✓ Real-time Feedback: Immediate responses allow iterative, agentic workflows
✓ Multi-step Operations: AI can chain multiple tool calls to accomplish complex tasks autonomously

How Garchi MCP Works

A seamless flow from AI request to content action

1

AI Makes Request

User instructs the AI to perform a task like 'Create a new page in Garchi CMS with 4 sections'

2

MCP Route

The request is routed through the MCP protocol to the appropriate Garchi CMS tool

3

Tool Execution

Your AI client executes the correct available tool with proper validation

4

Result Feedback

AI receives structured results and can continue with follow-up actions

You can also ask your AI client to troubleshoot your code of Garchi CMS integration or generate associated Garchi CMS code snippets for rendering your content.

Why use Garchi CMS MCP Server?

Connect your AI tools directly to your content management workflow

🤖

AI-Assisted Content Creation

Ask your AI client to create page structures, blog posts, or data entries directly in Garchi—no context switching between tools.

⚡

Direct CMS Access

Your AI can read current content states and make informed decisions about what to create based on your existing structure.

🎯

Skip the Repetitive Setup

Let AI handle the initial content scaffolding—creating sections, mapping fields, organizing categories—while you focus on the actual content and strategy.

📚

Full Content Control

Create pages with custom sections, manage blog articles with proper metadata, and organize content with categories and relationships—all through natural AI conversations.

🔗

Works with Your AI Client

Built on the Model Context Protocol standard. Use it with Claude Desktop, Cline, or any MCP-compatible tool you prefer.

Available Tools

A complete toolkit for managing your Garchi CMS through AI

📄 Page Management

Create Page Write

Create a new headless web page with title, description, and URL path.

→ Unique path validation

→ SEO-friendly descriptions

→ Auto-generated on creation

List Pages Read

Retrieve all headless pages in a space with metadata and timestamps.

→ Full page details

→ Creation timestamps

→ Latest updates info

Get Page Read

Fetch details of a specific page by slug.

→ Retrieve by slug

→ Includes all metadata and content

→ Fast lookups

🎨 Section Templates & Sections

Create Section Template Write

Define reusable section templates with configurable props that map to frontend components.

→ Multiple prop types (text, select, media, etc.)

→ Component mapping ready

→ Framework-agnostic

List Section Templates Read

Browse all available section templates in a space.

→ All templates with props

→ Ready for selection

→ Complete prop details

Create Section Write

Add a section instance to a page using a template.

→ Template-based creation

→ Supports nested sections

→ Auto-ordering

📝 Content Items

Create Data Item Write

Create blog articles, products, documents, or any itemable content type.

→ Rich HTML descriptions

→ Category associations

→ SKU & inventory tracking

List Data Items Read

Query all data items with filtering and pagination support.

→ Filter by category

→ Paginated results

→ Search capabilities

Create Meta for Item Write

Add additional properties to your content items.

→ Key value pair to extend your data item

→ Custom attributes

List Item Meta Read

Retrieve metadata for any content item.

→ All metadata fields

🏷️ Categories

Manage Category Write

Create and update content categories with full control.

→ Create new categories

→ Update existing ones

→ Rank management

List Categories Read

Retrieve all categories in a space for organization and selection.

→ Complete category list

→ Ranking info

→ Fast lookups

Installation & Integration Guide

Step-by-step instructions to connect your AI client to Garchi MCP Server

✅ Prerequisites

Before integrating Garchi MCP, ensure you have:

✓ A Garchi CMS account with at least one space created
✓ API keys generated for your account
✓ An AI client that supports the MCP protocol (Claude, ChatGPT, or custom agent)
✓ Basic understanding of your content structure (pages, sections, items)

Integrate with AI client

The MCP server endpoint is https://garchi.co.uk/mcp

Authentication works via Bearer token. Pass your API key in the Authorization header with each request.

Your API key is scoped to your account and spaces. You'll only be able to create and manage content in spaces you own.

Cursor IDE

You can manually install the server in your Cursor IDE or using a Cursor web link to set it up automatically.

Manual integration steps

Configure custom MCP servers with a JSON file. Paste the following in mcp.json of your Cursor IDE.

{
  "mcpServers": {
    "GarchiCMS": {
      "url": "https://garchi.co.uk/mcp",
      "headers": {
        "Authorization": "Bearer YourAPITokenHere"
      }
    }
  }
}

Automatic integration

Click the button below to add Garchi MCP server to your Cursor IDE automatically.

Claude desktop

To install GarchiCMS MCP in Claude desktop, follow the steps below:

Open Claude desktop and click on your profile icon --> Settings --> Developer.
Click on Edit config and then open claude_desktop_config.json file.
Add the following JSON snippet inside the "mcpServers" object:
```
{
  "mcpServers": {
    "GarchiCMS": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://garchi.co.uk/mcp",
        "--header",
        "Authorization: Bearer ${YOUR_TOKEN}"
      ],
      "env": {
        "YOUR_TOKEN": "Provide your API token here"
      }
    }
  }
}
```
When successfully installed you should see the MCP in your list of Local MCP servers in Connector section.

Claude CLI

To install GarchiCMS MCP in Claude CLI, run the following command in your terminal:

claude mcp add --transport http GarchiCMS https://garchi.co.uk/mcp --header "Authorization: Bearer your-api-token-here"

Lovable

You can add GarchiCMS MCP as personal connector in Lovable by following the steps below:

In Lovable click on your profile icon --> Settings --> Connectors --> New MCP server under Personal Connectors (MCP).
Enter server name as GarchiCMS
Enter server URL as https://garchi.co.uk/mcp
Under authentication select Bearer token and provide your Garchi CMS API key
Click on Add Server

v0

You can add GarchiCMS MCP in v0 by following the steps below:

Click Plus icon in v0 prompt input field --> MCPs --> Add MCP --> Custom MCP.
Enter server name as GarchiCMS
Enter server URL as https://garchi.co.uk/mcp
Under authentication select Bearer token and provide your Garchi CMS API key
Click on Add

VSCode

To install GarchiCMS MCP in VSCode, follow the steps below:

Open VSCode and open command palette using Ctrl+Shift+P (Windows/Linux) or Cmd+Shift+P (Mac).
In command palette type MCP: Open User Configuration
Once the mcp.json is opened, add this to your configuration of server:
```
{
  "servers": {
    "GarchiCMS": {
      "url": "https://garchi.co.uk/mcp",
      "type": "http",
      "headers": {
        "Authorization": "Bearer YourAPITokenHere"
      }
    }
  }
}
```
Once you save the file, VS Code detects the newly added GarchiCMS server. Sometimes you might need to restart VS Code for the changes to take effect.

Other clients

Garchi CMS can be used with other MCP client using your API key for authentication as Authorization Bearer token.

Add the following configurations in your MCP client

>=

{
  "mcpServers": {
    "GarchiCMS": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://garchi.co.uk/mcp",
        "--header",
        "Authorization: Bearer your-api-token-here"
      ]
    }
  }
}

💡 Examples

Here are some examples in which you could make your AI client interact with GarchiMCP server. Remember, these are just suggestions—feel free to experiment and find what works best for your workflow!

---> Create a landing page on Garchi CMS using 5 sections for space "Your space uid here"
---> List all the pages available in my Garchi CMS space "Your space uid here"
---> Create a blog post in Garchi CMS for category "Tech News" with title "Latest in AI". Generate a SEO-friendly body for the article and a URL slug.
---> Help me render the following Garchi CMS page structure in React.
---> And much more!...

Need Help?

Resources and support options

📖

API Reference

Complete documentation of all API endpoints, request/response formats, and error codes.

View Docs

💬

Contact us

We are there for you! Reach out to our support team for assistance with integration or troubleshooting.

Contact Support

Garchi MCP Server