Fireberry Apps MCP

Build Fireberry Apps with an AI assistant by connecting it to the Fireberry CLI through the Model Context Protocol (MCP).

Fireberry Apps MCP lets AI assistants such as Claude Code, Cursor, VS Code, and other MCP-compatible IDEs understand your app structure, scaffold components, and run Fireberry development workflows from natural-language requests.

What is Fireberry Apps MCP?

Fireberry Apps MCP exposes the Fireberry App development workflow to AI coding assistants. The Fireberry CLI acts as the bridge between your AI assistant and the Fireberry Apps platform.

Instead of manually running CLI commands, you can ask your assistant to:

• Create Fireberry Apps
• Generate components
• Build and deploy apps
• Install apps into your Fireberry account
• Read and update an app manifest
• Follow Fireberry best practices while generating code

What you can build

Use Fireberry Apps MCP to create each type of Fireberry App component.

Record components

Build custom experiences inside CRM records, such as:

• Customer Health Dashboard
• AI Lead Summary
• Related Data Viewer
• Financial Analysis Panel

Side menu apps

Create full applications inside Fireberry, such as:

• Analytics Dashboard
• Custom Reports
• Inventory Management
• Internal Operations Tools

Global menu apps

Create organization-wide applications, such as:

• KPI Dashboard
• AI Command Center
• Marketing Hub
• Executive Reporting

Prerequisites

Complete these steps before you connect an AI client:

1. Install Node.js 18 or later.
2. Install npm.
3. Create or sign in to your Fireberry account.
4. Create a Fireberry API token.
5. Install an MCP-compatible client, such as Claude Code, Cursor, or VS Code.

Install and authenticate the Fireberry CLI

1. Install the latest Fireberry CLI globally:

npm install -g @fireberry/cli@latest

2. Authenticate the CLI with your Fireberry account:

fireberry init

3. Verify that the CLI is installed:

fireberry --version

The CLI provides the commands you use to create, deploy, debug, and install Fireberry Apps.

Enable MCP

Configure your AI client to run the Fireberry CLI as an MCP server. Add the following configuration to your client's MCP settings:

{
  "mcpServers": {
    "fireberry-apps": {
      "command": "npx",
      "args": [
        "-y",
        "@fireberry/cli"
      ]
    }
  }
}

Use Fireberry Apps MCP

After you connect the MCP server, open your AI assistant and describe the app or change you want to make.

Create an app

Create a Fireberry App called Sales Dashboard.

Add a component

Add a record component for Contacts that displays AI-generated summaries.

Generate a UI

Create a side menu app that shows a chart of opportunities by stage.

Build and deploy

Build and deploy my app.

Install an app

Install the app into my Fireberry account.

Typical development workflow

1. Install the Fireberry CLI.
2. Connect your AI client to the MCP server.
3. Open Claude, Cursor, or another MCP-compatible client.
4. Describe the app you want to build.
5. Review the code your assistant generates.
6. Build the app.
7. Deploy the app.
8. Install it into your Fireberry account.

Behind the scenes

Your AI assistant uses the Fireberry CLI to run the same operations you can run manually.

Related documentation

• Create & Deploy an App
• Fireberry CLI
• Fireberry SDK
• Fireberry API MCP integration