Skip to main content

Integrations

MCP

Connect LoopMessage Documentation to AI Tools with MCP

LoopMessage provides a remote Model Context Protocol (MCP) server that allows compatible AI tools and coding assistants to access the latest LoopMessage documentation directly.

Instead of manually copying documentation into your AI assistant, you can connect the LoopMessage MCP server once and let your AI tool search the Helpdesk, find API documentation, and retrieve relevant integration details when needed.

MCP Server URL

Use the following URL to connect to the LoopMessage MCP server:

https://mcp.loopmessage.com/mcp

The server uses the Streamable HTTP MCP transport.

What can the LoopMessage MCP server do?

Once connected, your AI assistant can use LoopMessage documentation as an additional source of context.

Depending on the task, it can:

  • Search the LoopMessage Helpdesk.
  • Search the LoopMessage API documentation.
  • Retrieve the full documentation for a specific feature.
  • Find the appropriate API endpoint for a task.
  • Understand request parameters, headers, and payloads.
  • Generate API request examples for supported programming languages.
  • Help troubleshoot integrations using information from the LoopMessage documentation.

For example, you can ask your AI assistant:

How do I send an iMessage using the LoopMessage API?
Find the LoopMessage API endpoint for sending a message to an iMessage group.
Show me a Python example for sending a message with LoopMessage.
What parameters are required when sending a voice message?
How do LoopMessage webhooks work?

The AI assistant can search the connected LoopMessage documentation before answering.

How to connect

Open the MCP or integrations settings in your AI tool.

Add a new remote MCP server and use:

Name: LoopMessage Documentation
URL: https://mcp.loopmessage.com/mcp

If your client asks you to select a transport, choose:

Streamable HTTP

Save the configuration and enable the LoopMessage MCP server.

The exact location of MCP settings depends on the AI application you use. Look for sections named MCP, MCP Servers, Connectors, Integrations, or Developer Tools.

Example configuration

Some MCP clients allow remote servers to be configured using a JSON configuration.

A typical configuration may look similar to:

{
  "mcpServers": {
    "loopmessage": {
      "url": "https://mcp.loopmessage.com/mcp"
    }
  }
}

Configuration formats vary between MCP clients, so refer to your client’s documentation for the exact format it expects.

Using LoopMessage MCP in your coding workflow

After connecting the server, you can use it while building or maintaining a LoopMessage integration.

For example, you can ask your coding assistant:

I need to send an iMessage to a contact using LoopMessage.
Find the correct API endpoint and implement it in Python.

Or:

Using the LoopMessage documentation, create a Node.js function that sends
a message and handles the API response.

Or:

Check the LoopMessage documentation and tell me which webhook event I should
use for this feature.

This is particularly useful in AI-powered development environments because the assistant can retrieve LoopMessage-specific information instead of relying only on its built-in knowledge.

API code examples

The LoopMessage documentation supports API request examples for multiple programming languages and HTTP clients.

When available, the MCP server can provide examples generated from the same API request definitions used by our documentation website.

This helps keep examples consistent between the Helpdesk and AI-assisted development workflows.

Supported examples may include:

  • cURL
  • JavaScript / TypeScript
  • Node.js
  • Python
  • PHP
  • Ruby
  • Java
  • C#
  • Go

You can simply ask your AI assistant for the language you are using.

For example:

Show me the Python Requests example for this LoopMessage endpoint.

Always up-to-date documentation

The MCP server is connected to the same documentation source used by the LoopMessage website.

When LoopMessage documentation is updated and a new version of the Helpdesk is deployed, the MCP knowledge is updated as part of the same deployment process.

This means AI tools connected through MCP can use the current version of our published documentation without requiring you to download or manually update documentation files.

MCP vs. copying documentation manually

Without MCP, developers often need to:

  1. Find the relevant Helpdesk article.
  2. Copy the documentation.
  3. Paste it into an AI assistant.
  4. Repeat the process for additional API endpoints or features.

With MCP, the AI assistant can retrieve relevant LoopMessage documentation when it needs additional context.

This makes MCP especially useful when working with:

  • AI coding assistants
  • AI-powered IDEs
  • development agents
  • automated integration workflows
  • debugging and troubleshooting workflows

Troubleshooting

If your AI client cannot connect to the MCP server, verify that the configured URL is exactly:

https://mcp.loopmessage.com/mcp

Make sure your client supports remote MCP servers and the Streamable HTTP transport.

If the client supports MCP but only supports local stdio servers, it may not be able to connect directly to the LoopMessage remote MCP endpoint.

Security

The LoopMessage MCP server provides access to published LoopMessage documentation.

Connecting the documentation MCP server does not automatically give an AI assistant access to your LoopMessage account, API credentials, messages, or other private account data.

Never paste your LoopMessage API key into an AI conversation unless you understand how that AI tool handles sensitive information.

The LoopMessage MCP server uses a remote HTTP connection, so there is no package to download or local server to run.

Use the following MCP endpoint:

https://mcp.loopmessage.com/mcp

Below are example setup instructions for several popular development environments and AI assistants.

Visual Studio Code

Visual Studio Code supports remote MCP servers and can use their tools from AI Agent mode.

For a project-specific configuration, create a .vscode/mcp.json file in your project:

{
  "servers": {
    "loopmessage": {
      "type": "http",
      "url": "https://mcp.loopmessage.com/mcp"
    }
  }
}

After saving the file:

  1. Open the Chat view in Visual Studio Code.
  2. Switch the chat to Agent mode.
  3. Open the tools picker and make sure the LoopMessage MCP server and its tools are enabled.
  4. Ask the agent a LoopMessage-related question.

For example:

Using the LoopMessage documentation, show me how to send an iMessage in Python.

You can also configure MCP servers in your VS Code user profile if you want the LoopMessage documentation to be available across multiple projects. VS Code supports remote MCP servers over Streamable HTTP.

JetBrains IntelliJ IDEA

If you use JetBrains AI Assistant in IntelliJ IDEA, you can connect the LoopMessage MCP server directly from the IDE.

Open:

Settings → Tools → AI Assistant → Model Context Protocol (MCP)

Click Add, select the HTTP connection option, and use the following configuration:

{
  "mcpServers": {
    "loopmessage": {
      "url": "https://mcp.loopmessage.com/mcp"
    }
  }
}

You can choose whether the server should be available only for the current project or globally across your IDE.

Click OK, then Apply. Once the connection is established, the tools exposed by the LoopMessage MCP server become available to JetBrains AI Assistant.

For example:

Find the LoopMessage API endpoint for sending a message to an iMessage group and create a Java implementation.

JetBrains AI Assistant supports remote MCP servers using Streamable HTTP. Note that the AI Assistant → Model Context Protocol (MCP) settings are used to connect the IDE to external MCP servers; the separate MCP Server settings in JetBrains products are intended for exposing the IDE itself to external AI agents.

JetBrains PyCharm

PyCharm uses the same MCP integration provided by JetBrains AI Assistant.

Open:

Settings → Tools → AI Assistant → Model Context Protocol (MCP)

Click Add, choose the HTTP connection option, and enter:

{
  "mcpServers": {
    "loopmessage": {
      "url": "https://mcp.loopmessage.com/mcp"
    }
  }
}

Click OK and Apply.

Once connected, PyCharm AI Assistant can use the LoopMessage documentation when answering questions or helping you write integration code.

For example:

Using the LoopMessage documentation, implement sending a message with Python Requests.

You can configure the connection globally or only for the current project. JetBrains AI Assistant supports Streamable HTTP for remote MCP servers, so no local proxy is required for the LoopMessage endpoint.

ChatGPT

ChatGPT supports custom MCP-powered apps for supported workspace plans.

If your ChatGPT workspace provides access to custom MCP apps:

  1. Enable Developer mode if required by your workspace.
  2. Open Settings → Apps or the corresponding workspace Apps settings.
  3. Select Create to create a custom app.
  4. Enter the LoopMessage MCP endpoint:
https://mcp.loopmessage.com/mcp
  1. Complete the setup and enable the app for your account or workspace.

Because the LoopMessage documentation MCP is intended for public documentation, no LoopMessage account credentials are required simply to search the published documentation.

After it is enabled, you can ask ChatGPT questions such as:

Use the LoopMessage documentation to explain how delivery webhooks work.

or:

Find the correct LoopMessage endpoint for sending an iMessage and show me the Node.js implementation.

Direct creation and testing of custom MCP apps in ChatGPT currently depends on the ChatGPT plan and workspace permissions. OpenAI currently documents this functionality for Business and Enterprise/Edu workspaces on ChatGPT web, with administrator or developer permissions controlling who can add custom apps.

Codex

Codex can connect directly to remote MCP servers from the Codex CLI or Codex IDE extension.

The easiest way to add LoopMessage is:

codex mcp add loopmessage --url https://mcp.loopmessage.com/mcp

Verify that the server was added:

codex mcp list

Codex shares MCP configuration between the CLI and IDE extension, so you normally only need to configure the server once.

Alternatively, add the server manually to:

~/.codex/config.toml

with:

[mcp_servers.loopmessage]
url = "https://mcp.loopmessage.com/mcp"

You can then ask Codex:

Use the LoopMessage MCP documentation to find the API for sending a group message and implement it in this project.

If you want Codex to consistently consult LoopMessage documentation while working on your integration, you can also add an instruction to your project’s AGENTS.md, for example:

Always use the LoopMessage MCP server when you need information about the LoopMessage API or Helpdesk documentation.

OpenAI documents both the codex mcp add ... --url ... command and direct config.toml configuration for remote documentation MCP servers.

Claude

Claude supports remote MCP servers through custom connectors.

For individual accounts where custom connectors are available:

  1. Open Customize → Connectors.
  2. Click + and select Add custom connector.
  3. Enter a name such as:
LoopMessage Documentation
  1. Enter the MCP URL:
https://mcp.loopmessage.com/mcp
  1. Click Add.

After adding the connector, enable it for a conversation from the + → Connectors menu.

You can then ask Claude:

Search the LoopMessage documentation and explain how to send an iMessage with the API.

For Team and Enterprise organizations, an Owner or Primary Owner may first need to add the connector from:

Organization settings → Connectors → Add → Custom → Web

Team members can then connect to the configured LoopMessage connector from their own Customize → Connectors settings.

Claude’s remote custom connectors are available across Claude web and Claude Desktop, with availability and administration depending on the account plan. Anthropic currently supports remote custom connectors on Free, Pro, Max, Team, and Enterprise plans; Free accounts are limited to one custom connector.