# Amplience MCP for AI Agents MCP

> Amplience connects your content management system directly to your AI agent. This MCP lets you query complex content structures, pull metadata, update schemas, and push finished articles or product listings live—all through conversation.

## Overview
- **Category:** ecommerce
- **Price:** Free
- **Tags:** digital-asset-management, content-architecture, edge-delivery, schema-management, enterprise-content

## Description

Stop clicking through tabs just to manage content. Amplience links your headless CMS to any intelligent AI client, letting you treat enterprise content architecture like a chat session. You can tell the agent exactly what needs changing: 'Update the hero banner's call-to-action and push it live.' The system handles fetching the current structure, verifying schemas, making the necessary modifications, and deploying the changes directly to your edge delivery API. Because this MCP is hosted on Vinkius, you connect once from any compatible client—like Claude or Cursor—and get immediate access to managing content across your entire digital ecosystem. It’s about using natural language to perform complex publishing tasks.

## Tools

### create_content_item
Generates an entirely new, structured content piece that matches a defined schema within a specific folder.

### delete_content_item
Permanently removes a content item from the repository database after validating its version status.

### get_content_item
Retrieves all configuration data for a single content item, including its schema revision lock.

### list_content_items
Pulls paginated lists of content items from an entire repository for bulk review or auditing.

### get_delivery_content
Retrieves the exact structural data needed to verify what is currently being served by the Delivery CDN blocks.

### list_folders
Lists all organizational folders within a specific content repository.

### list_hubs
Shows every active workspace (Hub) you have access to in Amplience.

### list_repositories
Lists all content repositories that exist within a chosen Hub.

### publish_content_item
Pushes the latest, modified version of a specific content item directly to the live delivery CDN.

### update_content_item
Modifies an existing content item's data structure while ensuring it matches its current schema requirements.

## Prompt Examples

**Prompt:** 
```
What are all the active repositories in my main Hub? I need to know where the product data lives.
```

**Response:** 
```
**Hub: Main Ecommerce Portal**

*   🛒 Product Catalog (ID: `6k33i1`)
*   ✍️ Blog Content (ID: `6m75o9`)
*   🖼️ Marketing Banners (ID: `9q82y3`)

You can now ask me to list the folders within any of these for more detail.
```

**Prompt:** 
```
Pull the structural metadata and schema lock for the 'Spring 2024' banner item.
```

**Response:** 
```
**Content Item: Spring Banner (ID: `5tYv92`)**

*   Schema Lock Version: 5
*   Payload:
    ```json
    {
        "title": "Spring Sales!",
        "href": "/shop/spring"
    }
    ```
This means you can change the title or link, as long as they match this structure.
```

**Prompt:** 
```
I just edited the product data. Can you publish it to the live CDN now?
```

**Response:** 
```
Confirmed. Publishing content item '5tYv92' now.

✅ **Status:** Published successfully.
*   **Target:** Live Delivery CDN
*   **Timestamp:** 2026-06-12T14:30:00Z
*   **Verification:** The content is live and visible at the root domain.
```

## Capabilities

### List all Amplience Hubs and Repositories
See every accessible workspace and the main content repositories available to your agent.

### Get Content Item Details (Schema & Payload)
Retrieve a full copy of any single piece of content, including its current schema definition and metadata for review.

### Create or Modify Structured Content
Generate entirely new drafts or change existing data items using structured JSON payloads that match the required schemas.

### Manage Deletion of Assets
Permanently remove content items from the repository database after validating their version status.

### Publish Content to Live CDN
Execute the final step by pushing approved and modified content versions directly to your live delivery network.

## Use Cases

### Updating Global Sale Banners
An e-commerce specialist needs to change the Black Friday banner across three different product lines. They ask their agent to locate all relevant items, use `list_content_items` to confirm they found every instance, and then instruct the agent to execute an `update_content_item` for the new dates.

### Building a New Blog Series
A marketing manager needs 10 drafts for a technical blog series. They prompt their agent to use `create_content_item` ten times, feeding it structured JSON payloads that adhere to the official 'Blog Post' schema.

### Cleaning Up Old Content
The team realizes an old product line is deprecated. They ask the agent to search for all content items related to that line using `list_content_items`, and then execute `delete_content_item` on the confirmed list.

### Verifying Live Site Status
A developer suspects a recent product update isn't showing up correctly. They ask the agent to use `get_delivery_content` to pull the exact structural matching for the current live CDN blocks and verify the output.

## Benefits

- Eliminate manual content audits. Use `list_content_items` to pull paginated lists of articles or product blocks, instantly giving you a full inventory of your repository.
- Accelerate publishing cycles. Instead of copying data into the CMS UI, use `create_content_item` to generate structured drafts based on defined schemas with simple chat prompts.
- Maintain technical accuracy. When reviewing content, `get_content_item` pulls the precise schema lock and payload, guaranteeing your AI agent works with the current version rules.
- Control the live deployment. After drafting changes, use `publish_content_item` to push approved content directly to the CDN edge without manual website publishing steps.
- Build reliable developer workflows. Frontend teams can read a specific schema's output validation structure by running `get_content_item`, which allows them to scaffold accurate code interfaces.

## How It Works

The bottom line is: You talk to your AI client using plain English commands, and it executes complex CMS operations against Amplience for you.

1. Enable the Amplience connector within your client environment.
2. Provide your specific Amplience Hub Name along with a Personal Access Token.
3. Use natural language prompts in your AI client to guide the agent through content retrieval, modification, and final publishing.

## Frequently Asked Questions

**How can I use Amplience MCP to update my website content without logging into Amplience?**
You don't have to log in. Your AI agent handles all the connection details and API calls for you. You just tell it what needs changing, and the changes get pushed live automatically.

**Does Amplience MCP help me find out what content schemas I need to follow?**
Yes. The agent can pull the structural metadata (schema lock) for any item you name. This gives you the exact JSON structure and field rules required before you try to create or update anything.

**Can Amplience MCP manage content across multiple websites or hubs?**
The MCP is designed to read all accessible Hubs, so it can help you audit and manage content assets across various environments within your enterprise setup.

**If I edit a banner using Amplience MCP, will that change show up on my live site?**
Not until you tell the agent to publish it. You must use the publishing tool to push the changes from draft status out to the active delivery CDN.

**Can I generate brand new content drafts using Amplience MCP?**
Yes, you can instruct your AI client to create completely structured content items. You provide the data, and the agent ensures it adheres to the correct schema before generating a draft.