Skip to main content
Glama

Sociona MCP Server

by fav-devs
RAILWAY_DEPLOYMENT.md2.95 kB
# Railway Deployment for Sociona MCP Server ## Overview The Sociona MCP Server can be deployed as a separate Railway service that provides AI tools for Claude Desktop integration. ## Prerequisites 1. **Railway CLI installed**: `npm install -g @railway/cli` 2. **Logged into Railway**: `railway login` 3. **Sociona API Key**: Get from your Sociona Developer Portal ## Deployment Steps ### Step 1: Create Railway Project ```bash # Create new project for MCP server railway init --name "sociona-mcp-server" ``` ### Step 2: Set Environment Variables ```bash # Set required environment variables railway variables set NODE_ENV=production railway variables set SOCIONA_API_KEY="sk_live_your_api_key_here" railway variables set SOCIONA_API_BASE="https://api.sociona.com/api/v1" ``` ### Step 3: Deploy ```bash # Deploy the MCP server railway up --detach ``` ### Step 4: Get Service URL ```bash # Get the Railway domain for your MCP server railway domain ``` ## Alternative: Add to Existing Project If you want to add the MCP server to your existing PostPal project: ```bash # Link to existing project railway link # Add as new service railway add --name "mcp-server" # Set environment variables for the MCP service railway variables set --service mcp-server SOCIONA_API_KEY="sk_live_your_key_here" railway variables set --service mcp-server SOCIONA_API_BASE="https://api.sociona.com/api/v1" railway variables set --service mcp-server NODE_ENV=production # Deploy railway up --service mcp-server --detach ``` ## Configuration for Claude Desktop Once deployed, use the Railway domain in your Claude Desktop config: ```json { "mcpServers": { "sociona": { "command": "node", "args": ["/app/dist/index.js"], "env": { "SOCIONA_API_KEY": "sk_live_your_key_here", "SOCIONA_API_BASE": "https://api.sociona.com/api/v1" } } } } ``` **Note**: MCP servers communicate via stdio, not HTTP. The Railway service runs the MCP server process, and Claude Desktop connects to it locally. ## Health Check The MCP server doesn't expose HTTP endpoints, but you can verify it's running by checking the Railway logs: ```bash railway logs --service mcp-server ``` You should see: `"Sociona MCP server running on stdio"` ## Troubleshooting ### Common Issues 1. **"SOCIONA_API_KEY environment variable is required"** - Ensure the API key is set in Railway variables 2. **Connection timeout** - MCP servers need to stay running; check Railway service status 3. **Permission denied errors** - Verify your Sociona API key has the required scopes ### Logs and Monitoring ```bash # View recent logs railway logs --service mcp-server --lines 50 # View real-time logs railway logs --service mcp-server --follow ``` ## Cost Optimization MCP servers are lightweight and don't need: - PostgreSQL database - Redis cache - High memory allocation Use Railway's smallest instance size for cost efficiency.

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/fav-devs/sociona-mcp-server'

If you have feedback or need assistance with the MCP directory API, please join our Discord server