MCP Configuration

Complete guide to configure and use Caramelo AI's Model Context Protocol (MCP) with Claude Desktop, ChatGPT, and Server-Sent Events (SSE).

MCP Integration

The Model Context Protocol (MCP) by Caramelo AI is an open standard that allows you to connect Artificial Intelligence assistants (such as Claude Desktop, ChatGPT, and Gemini) directly to your project’s data securely and in real-time.

With this integration, the AI of your choice gains the ability to query and analyze metrics, leads, conversations, and the performance of your campaigns autonomously and securely.


🚀 How to Access the MCP Panel

  1. Access the Caramelo AI Admin Dashboard.
  2. Navigate to your project.
  3. In the sidebar menu, click on Integrations.
  4. In the Model Context Protocol (MCP) card, click on Manage.

On this page, you will find your project’s exclusive Connection URL and the list of active connections.


⚙️ Configuration Instructions by Client

Your project’s connection URL has the following default format: https://api.carameloai.com/mcp/projects/<PROJECT_ID>

🟣 Claude Desktop (Local/stdio Connection)

To make your Caramelo AI project data available in your local Claude Desktop, you can use the npx bridge provided by the official MCP SDK.

  1. Open or create the Claude Desktop configuration file on your operating system:
    • Windows: %APPDATA%\Claude\claude_desktop_config.json
    • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  2. Insert the following structure under the mcpServers key (replacing the URL with the one copied from your Dashboard):
{
  "mcpServers": {
    "caramelo-vendedor": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/sdk",
        "connect",
        "https://api.carameloai.com/mcp/projects/YOUR_PROJECT_ID"
      ]
    }
  }
}
  1. Restart Claude Desktop. A plug icon will appear in the chat indicating that the tools have been loaded and are ready for use.

🟢 ChatGPT (Custom GPTs / Connectors)

ChatGPT supports remote integration with automatic authorization flow (OAuth 2.1) and dynamic client registration.

  1. In ChatGPT, go to Explore GPTs -> Create GPT.
  2. In the Configure tab, scroll down to Actions and click Create new action.
  3. In the Authentication section, select OAuth and fill in the settings using the connection base URL obtained in the dashboard. ChatGPT will perform auto-registration (DCR) on our server.
  4. When the user interacts with the Custom GPT for the first time, they will be redirected to the Caramelo AI login and consent screen.
  5. After approval, the connection will be established and listed under Active Connections in your dashboard.

🔵 Gemini CLI / SSE-Compatible Clients

For CLI tools or custom clients using Server-Sent Events (SSE) transport:

  • Configure the client by pointing to your project’s HTTP connection URL.
  • The MCP client will receive a 401 Unauthorized challenge with the WWW-Authenticate header containing the consent screen information.
  • Follow the client instructions to complete browser authentication.

📊 Available Tools for the AI

Once connected, the AI assistant will be able to trigger the following tools when requested in the chat:

Tool NameDescriptionReturned Data
list_projectsLists all projects available in the account.Quick identifiers and project metadata.
get_chatbot_metricsReturns aggregated metrics for the project’s chatbot.Active conversations, average response time, and lead conversion.
get_live_metricsObtains real-time metrics of interactions and messages.Detailed activity history in the last few hours.
get_lead_funnelDisplays the distribution of leads across funnel stages.Leads split by qualification and sales stages.
read_conversationAllows the AI to view recent messages from a chat.Content of the selected conversation (requires consent).
get_daily_usageShows daily resource and token consumption.Project limits and API/token consumption in the period.

🔒 Security and Control

[!IMPORTANT] Data Isolation: Each connection is strictly bound to the selected project. The assistant cannot view data from other projects without explicit, new authorization.

[!TIP] Instant Revocation: If you lose access to a device or want to terminate an AI assistant’s connection, go to the Active Connections section on the MCP page in the dashboard and click Revoke next to the corresponding connection. The token will be invalidated immediately.

[!WARNING] Rate Limiting: To ensure platform stability, each project has a limit of 30 requests per minute to the MCP server.


🙋 Frequently Asked Questions (FAQ)

Does the connection token expire?

Yes, the real_time_access_token is temporary and used to establish the WebSocket/SSE connection. Every time the connection is restarted or the page is reloaded, a new access token will be required to keep your data secure.

How is my data protected?

We use end-to-end encryption and short-lived authentication tokens (JWT). Furthermore, no AI can access your data without explicit authorization performed via the Caramelo AI dashboard.

Can I use the same connection URL for multiple AIs?

Yes, you can register multiple clients on the same connection URL. Each active connection will be listed separately in your admin dashboard, where they can be managed individually.

What should I do if the connection fails?

Verify that your project identifier is correct in the configuration. Also, make sure that your local environment (such as Claude Desktop) has internet access to download the official MCP SDK via npx.