
Safari MCP Server
Provides AI assistants with Safari browser automation and developer tools access, enabling LLMs to control Safari, access console logs, monitor network activity, and perform browser automation tasks.
README
Safari MCP Server
A Model Context Protocol (MCP) server that provides AI assistants with Safari browser automation and developer tools access. This server enables LLMs to interact with Safari, access console logs, monitor network activity, and perform browser automation tasks.
Features
- 🚀 Safari Browser Automation: Start, control, and manage Safari sessions
- 🔍 Developer Tools Access: Get console logs, network logs, and performance metrics
- 📸 Screenshots: Capture page screenshots for visual analysis
- 🕵️ Element Inspection: Inspect DOM elements and their properties
- ⚡ JavaScript Execution: Run custom JavaScript in the browser context
- 📊 Performance Monitoring: Access timing metrics and performance data
- 🔧 Multiple Sessions: Manage multiple Safari automation sessions
Prerequisites
System Requirements
- macOS only (Safari and SafariDriver are only available on macOS)
- Node.js 18+
- Safari 10+ (comes with built-in WebDriver support)
Safari Setup
Before using this MCP server, you need to enable Safari's automation features:
-
Enable Developer Menu:
- Open Safari → Preferences → Advanced
- Check "Show Develop menu in menu bar"
-
Enable Remote Automation:
- In Safari's menu bar: Develop → Allow Remote Automation
-
Authorize SafariDriver:
sudo safaridriver --enable
Enter your admin password when prompted.
Installation
-
Clone or navigate to the project directory:
cd C:\Users\<USERNAME>\RiderProjects\AIPacketAnalyzer\safari-mcp-server
-
Install dependencies:
npm install
-
Build the project:
npm run build
Usage
Running the Server
npm start
The server will start and listen for MCP protocol messages on stdio.
Development Mode
npm run dev
Available Tools
Session Management
safari_start_session
- Start a new Safari automation sessionsafari_close_session
- Close a Safari sessionsafari_list_sessions
- List all active sessions
Navigation & Page Info
safari_navigate
- Navigate to a URLsafari_get_page_info
- Get current page URL and title
Developer Tools Access
safari_get_console_logs
- Get browser console logssafari_get_network_logs
- Get network activity logssafari_get_performance_metrics
- Get page performance metrics
Browser Interaction
safari_execute_script
- Execute JavaScript in the browsersafari_take_screenshot
- Capture page screenshotssafari_inspect_element
- Inspect DOM elements
Example Usage with Claude Desktop
Add this configuration to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"safari-devtools": {
"command": "node",
"args": ["C:/Users/<USERNAME>/RiderProjects/AIPacketAnalyzer/safari-mcp-server/build/index.js"]
}
}
}
Example Commands
Once connected to an MCP client like Claude Desktop, you can use commands like:
Start a Safari session with ID "main" and navigate to Google:
- Start session: safari_start_session with sessionId "main"
- Navigate: safari_navigate to "https://www.google.com"
- Take screenshot: safari_take_screenshot
- Get console logs: safari_get_console_logs
Configuration Options
When starting a session, you can specify options:
{
"sessionId": "my-session",
"options": {
"enableInspection": true, // Enable Web Inspector
"enableProfiling": true, // Enable timeline profiling
"usesTechnologyPreview": false // Use Safari Technology Preview
}
}
Limitations
- Single Session Limit: Safari only allows one WebDriver session at a time
- macOS Only: This server only works on macOS systems
- Safari-Specific: Unlike Chrome DevTools Protocol, Safari has limited debugging API access
- Network Logs: Network logging depends on Safari's performance logs and may have limitations
Development
Project Structure
safari-mcp-server/
├── src/
│ ├── index.ts # Main entry point
│ ├── safari-mcp-server.ts # MCP server implementation
│ ├── safari-driver.ts # Safari WebDriver manager
│ └── types.ts # TypeScript type definitions
├── build/ # Compiled JavaScript
├── package.json
├── tsconfig.json
└── README.md
Building
npm run build
Watching for Changes
npm run watch
Cleaning Build
npm run clean
Troubleshooting
Common Issues
-
"Session not found" errors:
- Make sure you've started a session before using other commands
- Check that Safari didn't crash or close unexpectedly
-
"Remote Automation not enabled":
- Verify Safari's Developer menu is enabled
- Check that "Allow Remote Automation" is enabled in the Develop menu
-
Permission errors:
- Run
sudo safaridriver --enable
and enter your password - Make sure Terminal has the necessary permissions
- Run
-
Safari not responding:
- Only one Safari automation session can be active at a time
- Close any existing automation sessions before starting new ones
Debug Mode
To see detailed logging, you can modify the server to include more verbose output or check Safari's Web Inspector while automation is running.
Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests if applicable
- Submit a pull request
License
MIT License - see LICENSE file for details.
Related Projects
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.

E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.