Documentation
Home Getting Started MCP Integration Security Detection Subscription & Billing API Reference Troubleshooting

Getting Started with MCP Sentinel

Sign Up
Welcome to MCP Sentinel!

Secure your AI agents with real-time monitoring and threat detection. Get started in minutes with our Free plan.

What is MCP Sentinel?

MCP Sentinel is a security monitoring platform designed specifically for AI agents using the Model Context Protocol (MCP). It provides real-time threat detection, security analysis, and automated alerting to help you identify suspicious activities before they become problems.

Key Features

  • Real-time Monitoring: Track all MCP tool operations as they happen
  • AI-Powered Detection: Advanced threat analysis using machine learning
  • Instant Alerts: Email and Slack notifications for suspicious activities
  • Usage Analytics: Detailed insights into your AI agent behavior
  • Easy Integration: Simple setup with existing MCP servers

Quick Start (5 Minutes)

Step 1: Create Account

Sign up for a free account to get started immediately.

  • 1,000 events per month
  • 1 MCP server connection
  • Email alerts included
Create Account
Step 2: Connect Server

Add your first MCP server for monitoring.

  • Simple webhook integration
  • Test connection instantly
  • Demo data available
Add Server
Step 3: Configure Alerts

Set up your notification preferences to stay informed about threats.

  • Email notifications
  • Slack integration
  • Severity thresholds
  • Business hours filtering
Configure Alerts

Core Concepts

MCP Events

An MCP event represents a single operation performed by an AI agent through the Model Context Protocol. This includes:

  • Tool calls (file operations, web requests, code execution)
  • User prompts and agent responses
  • Resource access attempts
  • Error conditions and exceptions

MCP Servers

MCP servers are the endpoints that provide tools and resources to AI agents. You can connect multiple servers to monitor different aspects of your AI system:

  • File system servers
  • Database connection servers
  • Web API servers
  • Custom tool servers

Threat Detection

Our AI-powered detection system analyzes each event for potential security threats:

  • Code Injection: SQL injection, command injection attempts
  • Data Exfiltration: Unusual data access patterns
  • Privilege Escalation: Attempts to gain elevated permissions
  • Suspicious Patterns: Anomalous behavior detection

Integration Guide

Webhook Integration

The easiest way to connect your MCP server is through webhooks:

Example Webhook URL:
https://your-domain.replit.app/api/mcp/webhook/YOUR_SERVER_ID

Manual Log Upload

For testing or one-time analysis, you can upload MCP logs directly:

  1. Go to Upload Logs
  2. Select your MCP log files
  3. Review the analysis results

Try the Demo

Want to see MCP Sentinel in action? We provide sample data and demo scenarios:

Demo Environment

Pre-configured servers with sample events to explore all features.

Set Up Demo
Interactive Testing

Test MCP tools directly and see how events are processed.

Test Tools

Next Steps

Once you've completed the quick start, explore these advanced features:

MCP Integration

Learn advanced server configuration and monitoring options.

Read Guide →
Security Configuration

Customize threat detection rules and alert settings.

Configure Security →
API Integration

Integrate MCP Sentinel with your existing tools and workflows.

View API Docs →
Need Help?

Check our troubleshooting guide or contact support if you run into any issues.