MCP Integration
AI Agent Control with MCP
Model Context Protocol (MCP) enables you to control avSign.lite digital signage using natural language commands through AI assistants like Claude, ChatGPT, and other AI agents.
What is Model Context Protocol?
MCP is an open standard that allows AI assistants to interact with applications through well-defined tools and commands. Instead of manually clicking through 9 steps to publish a video, you simply tell your AI assistant: “Play the product video now” – and it happens instantly.
Why Use MCP with Digital Signage?
The Problem
Traditional digital signage management involves repetitive, multi-step workflows:
- Upload content: Navigate to Resources → Upload → Tag → Save
- Create playlist: Navigate to Playlists → Create → Add items → Set durations → Save
- Publish: Navigate to Publish page → Select playlist → Choose terminals → Click publish
- Schedule: Navigate to Schedules → Create → Set time ranges → Set recurrence → Assign terminals
Result: 9-20 manual steps for simple tasks, taking 5-20 minutes per operation.
The Solution
With MCP integration:
YOU: "Play the product video on lobby displays now"
AI: ✓ Content published and playing
Time: 5 seconds
YOU: "Schedule breakfast menu 6am-11am, lunch menu 11am-5pm, dinner menu 5pm-10pm daily"
AI: ✓ Dynamic schedule created with 3 time-based items
Time: 30 seconds
Workflow reduction: 95% faster with MCP
Key Benefits
| Feature | Traditional UI | With MCP |
|---|---|---|
| Play content immediately | 9 steps, 5 minutes | 1 prompt, 5 seconds |
| Check what’s playing | 3 page navigations | Instant answer |
| Schedule time-based content | 15 steps, 20 minutes | 1 prompt, 30 seconds |
| Emergency announcements | 11 steps, 5 minutes | 1 prompt, 5 seconds |
How It Works
- Enable MCP Server in avSign.lite Settings
- Connect your AI agent to the MCP endpoint
- Issue natural language commands through your AI assistant
- AI agent calls appropriate tools to execute your request
- Instant feedback confirms the action
Available Tools
avSign.lite provides 9 MCP tools for complete digital signage control:
| Tool | What It Does |
|---|---|
content.create |
Upload images, videos, text, or HTML content |
content.list |
Browse all available content |
device.status |
Check device health and current playback |
playlist.create |
Build playlists with multiple content items |
playlist.list |
View all playlists |
playlist.publish |
Play playlist immediately on displays |
playlist.reset |
Stop playback and return to home screen |
schedule.create |
Create time-based schedules with recurrence |
schedule.list |
View all active schedules |
View complete tool reference →
Real-World Use Cases
Restaurant Dynamic Menu
Goal: Show different messages throughout the day
Morning (6-11am): "Good morning! Breakfast menu available until 11 AM"
Afternoon (11am-5pm): "Welcome! Check out our lunch specials"
Evening (5pm-10pm): "Happy hour starts at 5 PM - Special cocktails!"
Traditional approach: Create 3 playlists, create 3 schedules, set time ranges, assign terminals (~20 minutes)
With MCP: One natural language prompt to create the entire schedule (~30 seconds)
Retail Flash Sale
Goal: Immediately show urgent promotion
Command: “Show 50% off flash sale announcement on all displays now”
Result: AI creates content, publishes immediately, overrides current schedule
Corporate Emergency Alert
Goal: Display emergency evacuation notice
Command: “Show emergency evacuation notice on all displays immediately with high priority”
Result: Critical announcement visible within seconds across all screens
Getting Started
Ready to control your digital signage with AI?
- Quick Setup Guide - Enable MCP in 2 minutes
- Available Tools - Complete API reference
- Use Cases - Real-world examples
Technical Details
- Protocol: MCP 2025-06-18 over StreamableHTTP
- Authentication: Token-based (optional)
- Rate Limiting: 100 requests/minute per client
- Default Port: 3333
- Security: TLS/HTTPS recommended for production
Compatibility
MCP integration works with:
- ✅ Claude (Anthropic) - Native MCP support
- ✅ ChatGPT (OpenAI) - Via MCP client libraries
- ✅ Custom AI agents - Standard MCP protocol
- ✅ All avSign.lite platforms (Windows, Linux, Android, Fire TV)
Next Steps: Get started with MCP →