GoLogin Puppet MCP Bridge

HTTP bridge for exposing the GoLogin Puppet MCP server to remote Flowise instances.

Architecture

┌─────────────────┐
│ Flowise (Docker)│ <- Remote server
│  on Server B    │
└────────┬────────┘
         │ HTTP
         ↓
┌─────────────────┐
│  HTTP Bridge    │ <- This server (94.237.49.107:3001)
│   (Express)     │
└────────┬────────┘
         │ stdio
         ↓
┌─────────────────┐
│  GoLogin Puppet │ <- Local MCP server
│   MCP Server    │
└─────────────────┘

Quick Start

1. Install Dependencies

cd /tools/dev/gologin-puppet-mcp-bridge
npm install

2. Start the Bridge Server

npm start

Or run in background:

nohup npm start > bridge.log 2>&1 &

Or with PM2:

pm2 start server.js --name gologin-mcp-bridge
pm2 save

3. Test the Bridge

# Check health
curl http://localhost:3001/health

# List available tools
curl http://localhost:3001/tools

# Call a tool (example)
curl -X POST http://localhost:3001/tools/call \
  -H "Content-Type: application/json" \
  -d '{
    "name": "list_sessions",
    "arguments": {}
  }'

Using with Flowise

Option 1: Direct HTTP API (Custom Integration)

In your Flowise flow, use a Custom Tool or API node to call:

List Tools:

GET http://94.237.49.107:3001/tools

Call Tool:

POST http://94.237.49.107:3001/tools/call
Body: {
  "name": "start_browser",
  "arguments": {
    "apiKey": "your-api-key",
    "profileId": "your-profile-id"
  }
}

Option 2: MCP Client Wrapper (Recommended)

If your Flowise server can access this machine via HTTP, you can use the included client wrapper.

On your Flowise server, create this config in CustomMCP node:

{
  "command": "npx",
  "args": ["-y", "mcp-remote-proxy@latest"],
  "env": {
    "BRIDGE_URL": "http://94.237.49.107:3001"
  }
}

Or if you want to install the flowise-client.js on the Flowise server:

# On Flowise server
curl -o /tmp/flowise-mcp-client.js http://94.237.49.107:3001/client.js

# Then in Flowise CustomMCP:
{
  "command": "node",
  "args": ["/tmp/flowise-mcp-client.js"],
  "env": {
    "BRIDGE_URL": "http://94.237.49.107:3001"
  }
}

Option 3: Serve the Client Script

Add this endpoint to server.js to serve the client script:

app.get('/client.js', (req, res) => {
    res.sendFile(__dirname + '/flowise-client.js');
});

Then Flowise can download and execute it.

Configuration

Environment Variables

• PORT - Bridge server port (default: 3001)
• BRIDGE_URL - For the client script (default: http://94.237.49.107:3001)

Firewall

Make sure port 3001 is accessible from your Flowise server:

# Check if port is open
sudo ufw allow 3001/tcp

# Or with iptables
sudo iptables -A INPUT -p tcp --dport 3001 -j ACCEPT

Example Workflow in Flowise

1. Add Custom MCP node with bridge configuration
2. Select tools you want (start_browser, navigate, screenshot, etc.)
3. Connect to Agent (ReAct Agent, OpenAI Functions Agent, etc.)
4. Add Chat Model (ChatOpenAI, ChatAnthropic, etc.)
5. Test with prompt:

Use the GoLogin browser tools to:
1. Start a browser session with my credentials
2. Navigate to https://example.com
3. Get the page title
4. Take a screenshot
5. Close the browser

Monitoring

# View logs
tail -f bridge.log

# Or with PM2
pm2 logs gologin-mcp-bridge

# Check status
pm2 status

Troubleshooting

Bridge won't start

• Check if port 3001 is already in use: lsof -i :3001
• Check if MCP server path is correct
• Check logs for errors

Flowise can't connect

• Test with curl from Flowise server: curl http://94.237.49.107:3001/health
• Check firewall rules
• Check if bridge is running: pm2 status

Tools not loading

• Check bridge logs: pm2 logs gologin-mcp-bridge
• Test tools endpoint: curl http://localhost:3001/tools
• Restart bridge: pm2 restart gologin-mcp-bridge

Security Notes

⚠️ This bridge has no authentication!

For production use, add:

• API key authentication
• Rate limiting
• IP whitelisting
• HTTPS/TLS

Example with basic auth:

app.use((req, res, next) => {
    const apiKey = req.headers['x-api-key'];
    if (apiKey !== process.env.API_KEY) {
        return res.status(401).json({ error: 'Unauthorized' });
    }
    next();
});