jamf-mcp
Enables natural language interaction with Jamf Pro, Jamf Protect, and Jamf Security Cloud, allowing users to manage devices, security alerts, and configurations through Claude Desktop.
README
jamf-mcp
A Claude Desktop MCP (Model Context Protocol) server that gives Claude a natural-language interface to Jamf Pro, Jamf Protect, and Jamf Security Cloud.
Ask Claude things like:
- "Show me all Macs running macOS 13 or older"
- "Update the department for serial number C02XL0THJGH7 to Engineering"
- "List all high-severity Jamf Protect alerts from the last 24 hours"
- "Create a smart group for all devices with a CRITICAL risk score"
- "What API roles do we have configured?"
Features
| Product | Tools | Capabilities |
|---|---|---|
| Jamf Pro | 37 | Computers, mobile devices, users, smart/static groups, policies, configuration profiles, patch policies, App Installers, scripts, extension attributes, categories, departments, buildings, printers, prestages, API roles & integrations |
| Jamf Protect | 6 | Security alerts, analytics (detection rules), enrolled computers |
| Jamf Security Cloud | 2 | Device risk levels, risk overrides |
| Setup | 2 | Configuration status, setup instructions |
Zero credentials required to start — tools for unconfigured products return setup guidance instead of errors.
Requirements
- uv (recommended) or pip
- Python 3.11+
- Claude Desktop
Installation
git clone https://github.com/OtakuXavier/jamf-mcp.git
cd jamf-mcp
uv sync
Configuration
1. Set up credentials
Copy .env.example to .env and fill in your credentials:
cp .env.example .env
# Jamf Pro — OAuth (recommended)
JAMF_PRO_URL=https://yourorg.jamfcloud.com
JAMF_CLIENT_ID=your-client-id
JAMF_CLIENT_SECRET=your-client-secret
# Jamf Protect
JAMF_PROTECT_URL=https://yourorg.protect.jamfcloud.com
JAMF_PROTECT_CLIENT_ID=your-protect-client-id
JAMF_PROTECT_PASSWORD=your-protect-client-secret
# Jamf Security Cloud
JAMF_SECURITY_CLIENT_ID=your-security-client-id
JAMF_SECURITY_CLIENT_SECRET=your-security-client-secret
JAMF_SECURITY_REGION=us
You only need to configure the products you use — the others will be skipped gracefully.
2. Add to Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (see claude_desktop_config.example.json for the full template):
{
"mcpServers": {
"jamf": {
"command": "uv",
"args": ["run", "--directory", "/path/to/jamf-mcp", "jamf-mcp"],
"env": {
"JAMF_PRO_URL": "https://yourorg.jamfcloud.com",
"JAMF_CLIENT_ID": "your-client-id",
"JAMF_CLIENT_SECRET": "your-client-secret"
}
}
}
}
Restart Claude Desktop. You should see the Jamf tools available in the tool menu.
Getting Jamf Pro API Credentials
- Log in to Jamf Pro → Settings → System → API Roles and Clients
- Create an API Role with the privileges you need (e.g., Read Computers, Update Computers)
- Create an API Client, assign the role, enable it
- Copy the Client ID and generate a Client Secret
Or ask Claude to do it for you once you have admin credentials: "Create an API client with read-only access to computers and mobile devices"
Authentication
Jamf Pro supports two auth methods:
- OAuth2 (recommended): Set
JAMF_CLIENT_ID+JAMF_CLIENT_SECRET - Basic auth fallback: Set
JAMF_USERNAME+JAMF_PASSWORD
All three API clients cache bearer tokens in memory and refresh automatically 60 seconds before expiry.
Architecture
src/jamf_mcp/
├── server.py # FastMCP server, registers all tools
├── config.py # Env-var config dataclasses
├── response.py # Shared ok() / err() / not_configured() helpers
├── api/
│ ├── pro.py # Jamf Pro HTTP client (XML + JSON, token cache)
│ ├── protect.py # Jamf Protect GraphQL client
│ └── security.py # Jamf Security Cloud REST client
└── tools/
├── setup.py # jamf_get_setup_status, jamf_configure_help
├── pro/ # 37 Jamf Pro tools across 8 modules
├── protect/ # 6 Jamf Protect tools
└── security/ # 2 Jamf Security Cloud tools
All tools return a consistent envelope:
{ "success": true, "message": "Found 42 computers", "data": { ... } }
API Documentation
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.