What is Jupiterone Mcp

jupiterone-mcp is a Model Code Predictor (MCP) server designed for integration with JupiterOne, enabling enhanced code prediction capabilities.

Use cases

Use cases for jupiterone-mcp include automating code predictions in CI/CD pipelines, enhancing developer productivity by providing code suggestions, and integrating security best practices into the development workflow.

How to use

To use jupiterone-mcp, clone the repository, set up a Python virtual environment, install dependencies, configure the MCP server with your JupiterOne credentials, and integrate it into your projects by copying the necessary configuration files.

Key features

Key features of jupiterone-mcp include seamless integration with JupiterOne, customizable configuration options, and support for Python environments.

Where to use

jupiterone-mcp can be used in software development projects that require code prediction and analysis, particularly in environments utilizing JupiterOne for security and compliance.

JupiterOne MCP Server

A Model Context Protocol (MCP) server for LLMs that integrates with JupiterOne.

Setup Instructions

1. Clone the Repository

git clone https://github.com/JupiterOne/jupiterone-mcp.git
cd jupiterone-mcp

2. Create and Activate a Python Virtual Environment

# Create virtual environment
python -m venv .venv

# Activate on macOS/Linux
source .venv/bin/activate

# Activate on Windows
# .venv\Scripts\activate

3. Install Dependencies

pip install -r requirements.txt

4. Configure the MCP Server (Specifically for Cursor - instructions for other tools will vary)

Update the .cursor/mcp.json file with your JupiterOne credentials:

{
  "mcpServers": {
    "jupiterone": {
      "command": "/ABSOLUTE/PATH/TO/jupiterone-mcp/.venv/bin/python",
      "args": [
        "/ABSOLUTE/PATH/TO/jupiterone-mcp/mcp_server.py"
      ],
      "env": {
        "JUPITERONE_API_KEY": "your-jupiterone-api-key",
        "JUPITERONE_ACCOUNT_ID": "your-jupiterone-account-id",
        "JUPITERONE_REGION": "us"
      }
    }
  }
}

Make sure to:

• Replace /ABSOLUTE/PATH/TO/ with the actual full path to the repository
• Add your JupiterOne API key and account ID
• Set the correct region (default is “us”)

5. Use the MCP Server in Your Projects

To use this MCP server in another project:

1. Copy the .cursor directory to your target project:

   cp -r /path/to/jupiterone-mcp/.cursor /path/to/your/project/
   

2. If you already have a .cursor/mcp.json file in your target project, merge the contents:

You should get a prompt in the bottom left of cursor that a new MCP server has been discovered, and asking if you want to install it. Click yes.

Usage

Once configured, you can use JupiterOne queries directly within Cursor through the MCP interface. The server provides the run_j1_query tool that accepts JupiterOne Query Language (J1QL) queries.

Example usage in Cursor:

You can now run JupiterOne queries by asking Cursor something like:
"Can you show me all my AWS accounts in JupiterOne?"

Troubleshooting

• Ensure your virtual environment is activated when manually running the server
• Check that all paths in the .cursor/mcp.json file are absolute paths
• Verify your JupiterOne API key and account ID are correct
• Make sure the MCP server is running before using it in Cursor