MCP VMS: Bring Your Surveillance System Into AI Conversations

Stop context-switching between your AI assistant and surveillance software. This MCP server puts your entire CCTV system at your fingertips through natural language, turning security monitoring from a manual task into an intelligent conversation.

The Problem with Traditional Surveillance Workflows

You're managing multiple camera feeds, hunting through recordings for specific timeframes, and manually controlling PTZ cameras across different interfaces. When you need to investigate an incident or check system status, you're stuck juggling surveillance software, timestamps, and channel numbers.

MCP VMS changes that equation entirely.

What You Get

Direct Camera Control Through Conversation

"Show me camera 3 from yesterday at 2:30 PM"
"Move camera 5 to preset position 2" 
"Which cameras were recording between 9-11 AM?"

Your AI assistant becomes your surveillance operator, handling the tedious interface navigation while you focus on the actual security analysis.

Real-Time System Intelligence

• Query recording availability across all channels instantly
• Fetch live or historical images without opening the VMS interface
• Get connection and recording status for your entire camera network
• Control PTZ presets through simple commands

Seamless Integration Works with existing VMS infrastructure through the vmspy library. No need to rebuild your surveillance setup - just plug this MCP server into your workflow.

Real-World Applications

Security Incident Response When investigating an incident, ask your AI to pull footage from multiple cameras around a specific time, check which cameras were active, and move PTZ units to preset positions - all through conversation instead of clicking through multiple interfaces.

Facility Management "Check if all loading dock cameras are recording properly" or "Show me the parking lot from this morning at 8 AM" - get instant facility oversight without logging into separate systems.

Automated Monitoring Build AI-powered monitoring workflows that can automatically check camera status, retrieve specific footage, and even control camera positions based on events or schedules.

Technical Integration

The server connects to your VMS through a simple configuration:

vms_config = {
    'url': '127.0.0.1',
    'port': 3300,
    'access_id': 'admin',
    'access_pw': 'admin',
}

Add it to your Claude Desktop config and you're operational:

{
  "mcpServers": {
    "vms": {
      "command": "uv",
      "args": ["--directory", "path/to/mcp-vms", "run", "mcp_vms.py"]
    }
  }
}

Core Capabilities:

• Channel information and status retrieval
• Recording date/time queries
• Live and recorded image fetching
• Live stream and playback dialog control
• PTZ preset positioning
• Comprehensive error handling

Why This Matters

Instead of teaching your team to navigate complex VMS interfaces, they can ask natural questions and get immediate results. Security reviews become conversations. System monitoring becomes queries. Camera control becomes requests.

For developers building security applications, this MCP server becomes your surveillance API layer, accessible through any MCP-compatible AI system.

Requirements:

• Python 3.12+
• VMS server installation
• vmspy and Pillow dependencies

Installation: Clone the repo, install the VMS server from surveillance-logic.com, download the vmspy library, and configure your connection settings. The server handles the rest.

Ready to make your surveillance system conversational? Your security workflows will never feel the same.