trackmcp
Back to directory

DreamFactory MCP Server enables AI assistants like Claude to directly query your databases through DreamFactory's auto-generated REST APIs.

4 stars TypeScriptAI & Machine Learning Updated Oct 15, 2025

Documentation

DreamFactory MCP

A Model Context Protocol (MCP) server for DreamFactory integration with Claude.

Overview

DreamFactory is a secure, self-hosted enterprise data access platform that provides governed API access to any data source, connecting enterprise applications and on-prem LLMs with role-based access and identity passthrough.

Desktop Extension Installation (Claude Desktop Only)

The easiest way to start using DreamFactory's MCP Server is by going to this repository's releases and downloading the most recent Desktop Extension file. Desktop Extensions are a new file format created by Anthropic (the company behind Claude.ai). Once downloaded, double-click the DXT file to start installation. Note you'll need to install Claude Desktop before doing so. After starting installation you'll see this prompt:

Claude Desktop Extension Prompt

Click the `Install` button and you'll see:

Claude Desktop Extension Prompt #2

Click `Install` and if all goes well you'll be prompted to input a DreamFactory API key and service URL:

Configuring your DreamFactory MCP Server

If you don't have this information readily available, just click `Cancel`. You can add it later by clicking on the Extensions tab inside Settings and then clicking `Configure`:

Viewing your Extensions

In either case, before you can use the DreamFactory MCP Server you'll need to manually enable it by flipping the Disabled switch to Enabled:

Enabling the extension

Manual Installation

1. Clone this repository and then install dependencies:

code
cd df-mcp
   npm install

3. Build the project:

code
npm run build

Setting up Claude Desktop

1. Install Claude Desktop from https://claude.ai/download.

2. Open Claude Desktop and navigate to Settings.

3. In the Settings menu, go to the "Developer" section.

4. Click on "Edit Config" to modify Claude's configuration.

5. Add the following configuration to enable the DreamFactory MCP server:

json
{
       "mcpServers": {
           "add": {
               "command": "node",
               "args": [
                   "/Users/dreamfactory/Documents/df-mcp/build/index.js"
               ],
               "env": {
                   "DREAMFACTORY_URL": "https://example.dreamfactory.com/api/v2/",
                   "DREAMFACTORY_API_KEY": ""
               }
           }
       }
   }

6. Important: Modify the path in the `args` array to match your local installation path if different from `/Users/dreamfactory/Documents/df-mcp/`.

7. Add your DreamFactory URL and API key to the `DREAMFACTORY_URL` `DREAMFACTORY_API_KEY` environment variables as required.

8. Save the configuration and restart Claude Desktop.

Note About Claude Desktop, NodeJS and NVM

DreamFactory MCP requires Node 22.10.0 or newer. Even if you are running Node 22.10 or newer, you may encounter a confusing issue if you're running NVM (Node Version Manager) which hampers installation. Sometimes these issues are further complicated if you're additionally using other developer tools such as Laravel Herd which make additional changes to Node paths.

If the MCP server log indicates the wrong version of Node is being used by Claude Desktop, you can easily force it to use the desired version by creating a shell script that looks something like this:

code
#!/bin/zsh
source ~/.zshrc
exec /Applications/Herd/config/nvm/versions/node/v22.17.1/bin/node

Call the script `node-for-claude.sh` or similar, and place it in `/usr/local/bin`. Then make it executable:

code
$ chmod +x node-for-claude.sh

Finally, update the `claude_desktop_config.json` file to point to this script:

code
{
  "mcpServers": {
        "df-mcp": {
            "command": "/usr/local/bin/node-for-claude.sh",
            "args": [
                "/Users/wjgilmore/Software/df-mcp/build/index.js"
            ],
            "env": {
               "DREAMFACTORY_URL": "",
                "DREAMFACTORY_API_KEY": ""
            }
        }
  }
}

Usage

Once configured, the DreamFactory MCP server will be available to Claude Desktop. You can use DreamFactory's capabilities through Claude's interface.

Docker MCP Setup for Claude Desktop

Building the Docker Image

bash
docker build -t df-mcp .

Configure Claude Desktop

Add this configuration to your Claude Desktop config file:

Location:

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
  • Linux: ~/.config/Claude/claude_desktop_config.json

Configuration

bash
{
  "mcpServers": {
    "dreamfactory-mcp": {
      "command": "docker",
      "args": [
        "run",
        "--rm",
        "-i",
        "--env", "DREAMFACTORY_URL=https://your-dreamfactory-instance.com/api/v2/",
        "--env", "DREAMFACTORY_API_KEY=your-api-key-here",
        "dreamfactory-mcp:latest"
      ]
    }
  }
}

Restart Claude Desktop

After updating the configuration, restart Claude Desktop to load the MCP server.

Testing the Docker Container

Run the included test script to validate the Docker setup:

bash
./test-docker.sh

This will verify that the container builds correctly, responds to MCP commands, and all tools are properly registered.

Development

  • Source code is located in the `src` directory
  • Built files are in the `build` directory
  • Use `npm start` to run the built server directly

Security

Found a security issue? Email us at security@dreamfactory.com.

Frequently asked questions

What is df-mcp?

df-mcp is DreamFactory MCP Server enables AI assistants like Claude to directly query your databases through DreamFactory's auto-generated REST APIs.

How do I install df-mcp?

Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.

Is df-mcp open source?

Yes — it is hosted on GitHub at https://github.com/dreamfactorysoftware/df-mcp and has 4 stars.

Related MCP tools

Run your own MCP server? See who uses it and what to fix.

Measure it with TrackMCP