6.4 KiB
Multi-User Chrome MCP System
🎯 Overview
A complete multi-user system where multiple users can simultaneously use Chrome extensions with voice commands through LiveKit agents, with complete session isolation and manual agent management.
✨ Key Features
🔑 Unique User ID Generation
- Each Chrome extension generates a unique random user ID:
user_{timestamp}_{random} - User ID is consistent across all components (Chrome → Server → Agent → Back to Chrome)
- No authentication required - anonymous sessions with strong isolation
🤖 Automatic LiveKit Agent Spawning
- Remote server automatically starts a dedicated LiveKit agent for each Chrome extension
- Each agent runs in its own process with the user's unique ID
- Agents are automatically cleaned up when users disconnect
🏠 Session Isolation
- Each user gets their own LiveKit room:
mcp-chrome-user-{userId} - Voice commands are routed only to the correct user's Chrome extension
- Multiple users can work simultaneously without interference
🎤 Voice Command Routing
- LiveKit agents include user ID in MCP requests
- Remote server routes commands to the correct Chrome extension
- Complete isolation ensures commands never cross between users
🚀 Quick Start
1. Start the Remote Server
cd app/remote-server
npm install
npm start
2. Install Chrome Extension
- Load the extension in Chrome
- Open the popup and click "Connect to Remote Server"
- The extension will generate a unique user ID and connect
3. LiveKit Agent (Manual)
- LiveKit agents are no longer started automatically when Chrome extensions connect
- Agents should be started manually when voice functionality is needed
- When started, agents use the same user ID as the Chrome extension for proper session isolation
🔄 User Flow
1. Chrome Extension generates unique user ID
2. Connects to remote server with user ID
3. Remote server automatically spawns LiveKit agent with same user ID
4. User speaks voice commands to LiveKit agent
5. Commands are routed to correct Chrome extension based on user ID
6. Chrome extension executes commands and returns results
🧪 Testing
Complete Integration Test
node test-complete-integration.js
Tests the full flow with multiple users, voice commands, and session isolation.
Voice Command Routing Test
node test-voice-command-routing.js
Verifies that voice commands are routed to the correct Chrome extension.
Multi-User Connection Test
node test-multi-user-complete.js
Tests multiple Chrome extension connections and LiveKit agent spawning.
📊 Example Scenarios
Scenario 1: Multiple Users Searching
- User 1 says: "Open Google and search for pizza"
- LiveKit Agent 1 → Remote Server → Chrome Extension 1 → Google search for pizza
- User 2 says: "Navigate to Facebook"
- LiveKit Agent 2 → Remote Server → Chrome Extension 2 → Navigate to Facebook
Result: Both users work independently without interference.
Scenario 2: Session Isolation
- User 1 has 5 tabs open
- User 2 has 3 tabs open
- User 1 says: "Close all tabs"
- Only User 1's tabs are closed
- User 2's tabs remain untouched
Result: Perfect session isolation maintained.
🏗️ Architecture Components
Chrome Extension (app/chrome-extension/)
- Generates unique user ID
- Connects to remote server via WebSocket
- Executes voice commands received from LiveKit agent
Remote Server (app/remote-server/)
- SessionManager: Tracks user sessions and connections
- LiveKitAgentManager: Automatically spawns/manages LiveKit agents
- ConnectionRouter: Routes commands to correct Chrome extension
- ChromeTools: Handles MCP tool execution with user context
LiveKit Agent (agent-livekit/)
- Processes voice commands using OpenAI/Deepgram
- Includes user ID in all MCP requests for routing
- Connects to user-specific LiveKit room
🔧 Configuration
Environment Variables
# LiveKit Configuration
LIVEKIT_URL=ws://localhost:7880
LIVEKIT_API_KEY=devkey
LIVEKIT_API_SECRET=secret
# Remote Server
PORT=3001
HOST=0.0.0.0
User ID Format
user_{timestamp}_{random}
Example: user_1703123456_abc123def
LiveKit Room Names
mcp-chrome-user-{userId}
Example: mcp-chrome-user-user_1703123456_abc123def
📈 Monitoring
Session Status
- View active sessions in remote server logs
- Each session shows user ID, connection status, and LiveKit agent status
- Automatic cleanup of inactive sessions
LiveKit Agent Status
- Each agent logs its user ID and room name
- Agents automatically restart if Chrome extension reconnects
- Process monitoring and cleanup
🔒 Security & Isolation
- Anonymous Sessions: No authentication required, temporary user IDs
- Session Isolation: Each user's commands only affect their own browser
- Process Isolation: Each user gets a dedicated LiveKit agent process
- Network Isolation: Commands routed by user ID, no cross-contamination
📚 Documentation
MULTI_USER_SYSTEM_GUIDE.md- Complete usage guidedocs/MULTI_USER_CHROME_LIVEKIT_INTEGRATION.md- Technical architecture- Test files demonstrate complete system functionality
🎉 Success Criteria
✅ Multiple Chrome Extensions: Each user gets unique ID and session
✅ Automatic Agent Spawning: LiveKit agents start automatically for each user
✅ User ID Consistency: Same ID flows through Chrome → Server → Agent → Chrome
✅ Voice Command Routing: Commands reach correct user's Chrome extension
✅ Session Isolation: Users work independently without interference
✅ Comprehensive Testing: Full test suite validates all functionality
🚨 Troubleshooting
Issue: LiveKit Agent Not Starting
Solution: Check Python environment and dependencies in agent-livekit/
Issue: Voice Commands Going to Wrong User
Solution: Verify user ID consistency in logs across all components
Issue: Chrome Extension Not Connecting
Solution: Ensure remote server is running on localhost:3001
Issue: Multiple Users Interfering
Solution: Check that each user has unique user ID and separate LiveKit room
🎤 Ready to experience multi-user voice automation? Start the system and connect multiple Chrome extensions to see the magic happen!